[Senate Hearing 117-148]
[From the U.S. Government Publishing Office]
S. Hrg. 117-148
EMERGING TECHNOLOGIES AND THEIR IMPACT
ON NATIONAL SECURITY
=======================================================================
HEARING
BEFORE THE
COMMITTEE ON ARMED SERVICES
UNITED STATES SENATE
ONE HUNDRED SEVENTEENTH CONGRESS
FIRST SESSION
__________
FEBRUARY 23, 2021
__________
Printed for the use of the Committee on Armed Services
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Available via: http://www.govinfo.gov
__________
U.S. GOVERNMENT PUBLISHING OFFICE
45-695 PDF WASHINGTON : 2022
-----------------------------------------------------------------------------------
COMMITTEE ON ARMED SERVICES
JACK REED, Rhode Island, Chairman JAMES M. INHOFE, Oklahoma
JEANNE SHAHEEN, New Hampshire ROGER F. WICKER, Mississippi
KIRSTEN E. GILLIBRAND, New York DEB FISCHER, Nebraska
RICHARD BLUMENTHAL, Connecticut TOM COTTON, Arkansas
MAZIE K. HIRONO, Hawaii MIKE ROUNDS, South Dakota
TIM KAINE, Virginia JONI ERNST, Iowa
ANGUS S. KING, Jr., Maine THOM TILLIS, North Carolina
ELIZABETH WARREN, Massachusetts DAN SULLIVAN, Alaska
GARY C. PETERS, Michigan KEVIN CRAMER, North Dakota
JOE MANCHIN III, West Virginia RICK SCOTT, Florida
TAMMY DUCKWORTH, Illinois MARSHA BLACKBURN, Tennessee
JACKY ROSEN, Nevada JOSH HAWLEY, Missouri
MARK KELLY, Arizona TOMMY TUBERVILLE, Alabama
Elizabeth L. King, Staff Director
John D. Wason, Minority Staff Director
(ii)
C O N T E N T S
_________________________________________________________________
February 23, 2021
Page
Emerging Technologies and Their Impact on National Security...... 1
Members Statements
Statement of Senator Jack Reed................................... 1
The Prepared Statement of Senator James Inhofe................... 3
Witnesses Statements
Schmidt, Dr. Eric E., Co-Founder, Schmidt Futures................ 3
Smith, Mr. Brad L., President, Microsoft Corporation............. 13
Carlisle, General Herbert J., USAF (Ret.), President and Chief 28
Executive Officer, National Defense Industrial Association.
(iii)
EMERGING TECHNOLOGIES AND THEIR IMPACT ON NATIONAL SECURITY
----------
TUESDAY, FEBRUARY 23, 2021
United States Senate,
Committee on Armed Services,
Washington, DC.
The Committee met, pursuant to notice, at 9:32 a.m. in Room
SD-106, Dirksen Senate Office Building, Senator Jack Reed
(Chairman of the Committee) presiding.
Committee Members present: Senators Reed, Shaheen,
Gillibrand, Blumenthal, Hirono, Kaine, King, Warren, Peters,
Manchin, Duckworth, Rosen, Kelly, Inhofe, Wicker, Fischer,
Cotton, Rounds, Ernst, Tillis, Sullivan, Cramer, Scott,
Blackburn, Hawley, and Tuberville.
OPENING STATEMENT OF SENATOR JACK REED
Chairman Reed. I will call the hearing to order, and good
morning. And since this is the first open hearing since the
Senate has organized I would like to begin by once more
welcoming the new members of the committee, Senators Rosen,
Kelly, and Tuberville. We all look forward to working with you
this year, as we provide oversight to the Department of Defense
and craft the Fiscal Year 2022 National Defense Authorization
Act.
This morning the committee meets to examine the impact of
emerging technologies on national security. I want to thank the
three extremely well-qualified witnesses who are joining us
today to help us better understand this issue. Dr. Eric Schmidt
is the former Chief Executive Officer (CEO) of Google and chair
of the Defense Innovation Board, and currently co-chairs the
National Security Commission on Artificial Intelligence, which
was established by this committee. Mr. Brad Smith is the
president of Microsoft Corporation, and retired General
``Hawk'' Carlisle is the president and CEO of the National
Defense Industrial Association.
Each of you has unique and extensive technical, commercial,
and defense experience at the intersection of advanced
technology and the military that will help inform our
discussion. It is my hope that today we can begin to address a
number of key questions relating to emerging technologies and
national security, including what are the key emerging
technology areas and trends that will shape national security
and economic prosperity in the future; what actions could
accelerate or slow the operational use of these technologies;
how do you assess the standing of the United States in the
global competition to develop and deploy these emerging
technologies; and what specific recommendations do you have for
actions in policy, programs, or organizational reform that this
committee or the Pentagon should pursue to improve our ability
to deploy these technologies for national security.
The future national security environment will likely be
shaped by emerging technologies such as quantum computing,
biotechnology, hypersonics, 5G, and artificial intelligence. I
am concerned that the Defense Department is not postured
correctly to invest in the correct emerging technologies or to
play the appropriate role of co-developer and early adopter of
the advanced capabilities they will enable.
The technology development environment has become
globalized and extremely fast moving. We need to make sure that
we are looking at the right technologies, have the processes in
place to take advantage of them, and deliver new capabilities
to warfighters at the speed of technological change, and
faster, much faster, than our peer adversaries. Overlaying this
is the competition with China in both the national security and
economic sectors and their aggressive attempts to undercut our
current technological superiority.
We must also be concerned about the strength of our
national research and innovation enterprise, including the
workforce, the health of the manufacturing and industrial base,
and the infrastructure that we need to support technology
development.
Finally, all of this must be in light of budget constraints
and competing challenges for the Department of Defense (DOD),
namely balancing modernization with near-term readiness and
force structure. We also want to make sure that we are making
the best use of the great advantages that this nation possesses
in the global competition. For example, we have the world's
best innovators in defense industry and the commercial sector.
Are there ways that we can help them work more closely together
to produce next-generation defense systems.
We have the world's leading research universities, whose
efforts have led to all the emerging technologies we are
discussing today and also many of the technologies that we use
in our current force and even our daily lives. Are we still
making best use of their talents to support national security?
We are still the magnet for the world's best and brightest
technical minds. Are we positioning ourselves to continue to
attract that talent and to get them to work on the complex
national security challenges of the future?
The technologies and systems that we take for granted for
both national security, such as precision weapons, the nuclear
deterrent GPS [global positioning system], and the internet,
were all called emerging technologies at some point. It took
focused investment of resources and the time and toil of
countless scientific experts to solve the technical challenges
that inevitably occurred, but it also took leaders that were
willing to patiently protect those resources and people,
encourage risk-taking, and to accept and drive the changes
necessary to cut through the red tape and support these systems
moving from the lab into our operations. With today's emerging
technologies and changing world, we are faced with similar
decisions and challenges, and we need to ensure that we have
the same experts and leadership for success.
Again, I want to thank you all for your willingness to
appear today, and I look forward to your testimony.
Senator Inhofe is delayed, and he indicated that he would
prefer to have his statement submitted for the record. I ask
that that be submitted, without objection. So ordered. Thank
you very much.
[The prepared statement of Senator James Inhofe follows:]
prepared statement by senator james inhofe
Thank you, Mr. Chairman.
Dr. Schmidt, Mr. Smith and General Carlisle. Thank you for
being here to talk about this important topic today.
As highlighted in the 2018 National Defense Strategy
Commission report, the United States must stay ahead in several
emerging technologies to maintain or regain a warfighting
advantage against China and Russia.
Some of those key technologies are Artificial Intelligence
(AI), 5G, Hypersonics, Emerging Biotechnologies and Quantum
Computing, and Directed Energy.
The COVID-19 pandemic has shown that Department of Defense
emerging biotechnology research, including that done over the
last decade at the Defense Advanced Research Projects Agency
(known as DARPA) has been critical in fostering new
technologies that have been critical in vaccine and therapeutic
development in this pandemic.
I am looking forward to hearing our witnesses talk today
about what technologies are most important for us to adopt for
our warfighters to be best prepared for the future.
Russia and China are aggressively developing these
capabilities, and, in some cases, we are already behind or
falling behind. Without action, the United States may find
itself at a technological disadvantage in future conflicts.
However, the challenges facing the Department of Defense
are not just about developing new technologies, but the
Department must also reform its processes, policies and culture
to be able to more quickly adopt and deploy new technologies-
all while making sure that we are balancing the need to
modernize our military capabilities with maintaining near-term
readiness.
Thank you for appearing, and I look forward to your
testimony.
Chairman Reed. And now I will ask the witnesses to begin.
Dr. Schmidt, please.
STATEMENT OF DR. ERIC E. SCHMIDT, CO-FOUNDER, SCHMIDT FUTURES
Dr. Schmidt. Thank Mr. Chairman, I think I can speak for
all of us that we are incredibly proud to have been invited
here, and it is a great honor and privilege to be part of your
discussions.
I am one of these people who, like everyone in the room,
believes very strongly that America is a great country and that
our leadership is very, very important. I also believe that our
national security in the United States is tied to both our
economic security and our military security. And I am worried
that we do not understand the competitive threat from China to
what we are trying to do, and I want to take you through some
of the things that are going on.
In each of the following strategic areas, China is pushing
to meet or beat the work of the United States: semiconductors,
where both countries are dependent on Taiwan and South Korea;
artificial intelligence (AI), China catching up relatively
soon, according to their doctrine; energy: they are way down
the maturation curve, and we need to jump forward or lose that
industry; quantum: they have a well-funded effort and there are
important national security consequences from the use of
quantum in a number of areas; communications: we are all
familiar with the dominance of Huawei and the issues for
national security that is provided. You can see that the
success of Huawei in the developing world will be a long-term
problem for our country; and synthetic biology, the building of
life. China is busy building a biobank and is trying to sort of
come to global domination in a number of key areas.
These are contests of values as well as investments, and it
is important that American values, the things that we hold and
cherish so deep, are the winners in all of these technological
areas. We need to do a whole bunch of things, including focus
on advanced production, which covers manufacturing,
architecture, and assembly, and intelligence-augmented
infrastructure, everything from our roads and bridges to
pipelines to electric networks. This is how America wins.
So what we need to do is recognize that China is a very
significant competitor and that we need to respond to the sort
of things they are doing and make sure we stay well ahead. So I
will give you a set of examples, which will inform the
discussion.
The United States national security apparatus, and in
particular the DOD, treats software as a very low priority. It
needs to be treated as a very high priority. Software is going
to drive pretty much all of the interesting accomplishments in
the national security sense in the next 10 or 20 years, and
hiring and training and personnel policies that are similar to
the software companies are important.
We need to build missiles the way we now build cars. It
turns out that the modern car plan designs everything in a
design studio, knows everything, presses a button, and boom,
all that come out, and they work really, really well. The
bespoke design approaches, where the contractors today and the
primes operate, are completely counter to the way a Silicon
Valley company would operate. You put a design team together,
they figure it all out, they work very quickly, very much like
the original Lockheed Skunk Works. We have lost that, and it is
important to retain that.
We must make sure, for our economic strength, that the next
generation of technologies in AI, semiconductors, and so forth,
are successful not just for our commercial operations but our
national security.
If I continue to give you a few more examples, we are going
to have to have some kind of leadership out of the White House.
I am the chairman of the National Security Commission on AI.
Thank you. You all asked for it. It is coming out March 1. One
of its many recommendations is that there be a technology
competitiveness council at the White House, driven by the Vice
President, to get the kind of right attention on all of these
issues.
We are going to have to basically fund an AI research
network, one of our recommendations. We are going to have fund
biology labs, where you can order up the kind of biology that
you need and it shows up the next day, so you can continue to
be innovating. We are going to need to welcome high-skills
immigrants into the U.S., and keep our foreign-born PhDs here
in the country.
We are going to need a solution to the 5G problem. China
will soon have 1 billion people connected to a 1 gigabit
network on each of their phones. The United States strategy
does not have enough bandwidth allocated for 5G, and the
telecommunications companies just spent $80 billion to purchase
frequency in the C-band. That $80 billion went to the U.S.
Government. In my view, instead of spending it, to the United
States Government, it should have been used to spend to build
the infrastructure to build the 5G infrastructure to compete
with China and to provide leadership for us.
The important thing here, and I will finish up by saying,
is that the private sector is America's great strength. We move
faster and globally than any government could. Fast, iterative
design and product cycles are the key to competitiveness, and
we need global platforms or we will be forced to use the
Chinese ones, which is a disaster. I propose the combination of
what I said, adopt the AI Commission recommendations, which are
coming out on March 1, target the military systems that can be
accelerated by some of these new design approach--you are
wasting money with the existing design cycles. It is not
helping with preparedness. And then figure out a way to build
agreements between American industry--and, Mr. Chairman, you
already talked about this--and the military, and also build
very tight relationships with our trusted strategic partners in
other countries.
Thank you.
[The prepared statement of Dr. Schmidt follows:]
Prepared Statement by Dr. Eric E. Schmidt
Chairman Reed, Ranking Member Inhofe, Members of the Committee,
thank you for the opportunity to testify on the importance of emerging
technologies for the future of our national security.
I will begin with a broad view of the state of U.S. technology
leadership, then discuss the future defense landscape, and conclude
with some recommendations for the Pentagon.
I offer these views in my personal capacity, but they are informed
by my experience leading the National Security Commission on Artificial
Intelligence (NSCAI) and the Defense Innovation Board (DIB), as well as
my work in philanthropy, with Schmidt Futures, and in the private
sector. Many of my points here preview the conclusions and
recommendations in the AI Commission's forthcoming Final Report set to
be released publicly on March 1.
My argument today is straightforward: When it comes to emerging
technologies, our government needs to get the fundamentals right. I
mean that in two ways. First, to preserve national competitiveness, we
need to focus on the fundamental technologies that will have broad
impacts on our economy, our society, and our security. Second, to shape
the military we will need to defend the United States in the future, we
have to put the fundamental building blocks into place as soon as
possible. Those include the people, the research, the technology
infrastructure, and other basic elements that I will describe.
The AI Commission's Final Report includes many critical
recommendations to win the global technology competition and strengthen
national defense. I urge the Committee to seriously consider adopting
all of the recommendations that are relevant to your work, and also to
encourage your colleagues on other committees to do the same.
The logic for action is compelling.
global technology leadership and national security
Extending our global leadership position in technology is both an
economic and a national security imperative. Innovation is the
foundation of our economy, and the source of our military advantage.
Leadership gives our government and military access to the most
advanced available technologies. It puts us in the best position to
secure them against vulnerabilities. And it enables us to set standards
for their responsible use.
I am convinced that the threat of Chinese leadership in key
technology areas is a national crisis and needs to be dealt with
directly, now. The President had it exactly right in his speech in
Munich: the United States is in a ``long-term strategic competition
with China.''
China is pursuing technology leadership through strategic
investments in a wide range of critical technology areas, including
through the Made in China 2025 initiative. Consider artificial
intelligence, which is the fulcrum of this broader technology
competition. AI will be leveraged to advance all dimensions of national
power--from healthcare to food production to environmental
sustainability. The successful adoption of AI in adjacent fields and
technologies will drive economies, shape societies, and determine which
states exert influence and exercise power in the world. Many countries
have national AI strategies. But only the United States and China have
the resources, commercial might, talent pool, and innovation ecosystem
to lead the world in AI. In some areas of research and applications,
China is an AI peer, and it is already more technically advanced in
certain applications. Within the next decade, China could surpass the
United States as the world's AI superpower.
In addition to AI, China is seeking to lead the world in quantum
computing, fifth generation (5G) networks, and synthetic biotechnology,
among other areas. Beijing sees its national strategies in these areas
as mutually reinforcing. The CCP has made clear which technologies it
views as top national priorities. In each of these areas China is
pushing to meet or beat our work.
If China takes the lead, the first-mover advantages in developing
and deploying new technologies will make it difficult for the United
States to catch up. In critical sectors with strong network effects
like telecommunications, a winner-take-all dynamic raises the stakes
for rapidly developing leading technology platforms. The United States
Government must develop a unified strategy to advance and protect the
technologies that will underpin national competitiveness in the middle
decades of the 21st Century, even as we continue to cooperate with
competitors like China in areas of mutual interest.
a white house approach to national competitiveness in critical
technologies
The United States needs an integrated approach to federal
investments and policies across a range of emerging technologies. A
comprehensive national strategy would set and reinforce priorities and
would reconcile budget tradeoffs. The strategy should be led by the
White House. I strongly endorse the AI Commission's recommendation to
establish a new White House-led Technology Competitiveness Council.
This would be chaired by the Vice President and overseen by a senior
White House coordinator to ensure the President has the organization in
place to develop, drive, and fund a real national technology strategy.
A national strategy should focus on fundamental technologies with
broad impact on national competitiveness and security. A priority
shortlist should include AI, 5G, microelectronics, biotechnology, and
quantum computing. The importance of these areas is widely recognized.
The shortlist should also include advanced production (which covers
manufacture, agriculture, and assembly), as well as infrastructure
augmented by machine intelligence (everything from roads to bridges to
pipelines to electric networks).
Advanced production is essential to enable the country to produce
the goods it needs in the face of supply chain shocks, natural
disasters, epidemics, and so on. And it can permit leapfrogging through
greater efficiencies and energy optimization while reducing decaying
stockpiles of goods. The capacity to produce high-tech goods
domestically is critical to national security, both to maintain access
to finished goods and as a driver of innovation. The United States must
strive for self-reliance in industries that are critical to national
security or that would take too long to regenerate in the event of
protracted conflict.
New infrastructure is essential to handle emergencies (for example,
think of Texas's frozen gas supplies, or California's shifting
wildfires), permit tradeoffs among different modalities (trains versus
trucks versus pipelines), and reduce both environmental impact and
total cost of ownership. U.S. physical infrastructure remains largely
disconnected: no U.S. cities are ranked among the world's top 10 in
smart city connectedness, and only one is in the top 30. \1\ Maximizing
citizens' access to the digital economy, and more closely connecting
the physical and digital worlds, will be necessary to fuel future
growth. This can add a significant boost to national GDP.
---------------------------------------------------------------------------
\1\ Smart City Index, IMD, 8 (Oct. 2019), https://www.imd.org/
research-knowledge/reports/imd-smart-city-index-2019/
#:8:text=The%20Top%2010%20smartest%20cities,and%20Dusseldorf%20 (10th).
---------------------------------------------------------------------------
a more assertive government role
On a level playing field, the United States is capable of out
innovating any competitor. However, today, there is a fundamental
difference in approaches to innovation between the United States and
China that puts American leadership in peril. For decades, the United
States innovation model has been the envy of the world. The open
exchange of ideas and free markets, with targeted government
involvement to support basic research, are pillars of the American way
of innovation and reflect American values. In America, tech firms
compete for market share; they are not instruments of state power.
Most technology advances in the United States will be driven by the
private sector and universities. We must not lose an innovation culture
that is bottom-up, and infused with a garage startup mentality.
However, keeping things exactly the same as we have in the past is not
a winning strategy. Large tech firms cannot be expected to compete with
the resources of China or make the big, nation-wide investments the
United States will need to stay ahead. We will need a hybrid approach
that more tightly aligns government and private sector efforts to win.
The private sector is America's great strength; companies move
faster and more globally than any government could. However, given the
changing landscape, the U.S. Government must take a hands-on approach
to national technology competitiveness. Promoting a diverse and
resilient research and development (R&D) ecosystem and commercial
sector is a government responsibility. Expanding talent pipelines, more
quickly reforming immigration and visa authorities like H-1B to attract
the world's best, and improving our education system are all public
policy choices. Protecting critical intellectual property and thwarting
the systemic campaign of illicit knowledge transfer being conducted by
competitors is a government obligation. Protecting hardware advantages
and building resiliency into supply chains necessitates legislation and
federal incentives. Bringing together like-minded allies and partners
requires U.S.-led diplomacy.
democratizing ai research: a national research resource
Here is one concrete example of government action that could spur
nation-wide technology advances with benefits for overall national
competitiveness. Today, I worry that only a few big companies and
powerful states will have the resources to make the biggest AI
breakthroughs. Despite the diffusion of open source tools, the needs
for computing power and troves of data to improve algorithms are
soaring at the cutting edge of innovation. The government should
democratize access to compute environments, data, and testing
facilities in order to provide researchers beyond leading industry
players and elite universities the ability to pursue progress on the
cutting edge of AI. It can do this by creating a National AI Research
Resource (NAIRR), which would provide verified researchers and students
subsidized access to scalable compute resources, co-located with AI-
ready government and non-government data sets, educational tools, and
user support. \2\ It should be created as a public-private partnership,
leveraging a federation of cloud platforms. \3\ The AI Commission has
detailed plans to implement this recommendation.
---------------------------------------------------------------------------
\2\ Acting on a recommendation NSCAI issued in our First Quarter
Recommendations, Congress has taken the first step to establish the
NAIRR in the Fiscal Year 2021 National Defense Authorization Act,
creating a task force to develop a roadmap for a future NAIRR. The
result of this effort will be due to Congress 18 months after
appointment of task force members. See Pub. L. 116-283, William M.
(Mac) Thornberry National Defense Authorization Act for Fiscal Year
2021, 134 Stat. 3388 (2021); see also First Quarter Recommendations,
NSCAI at 2 (Mar. 2020), https://www.nscai.gov/previous-reports/.
---------------------------------------------------------------------------
digital infrastructure: getting 5g right
Promoting the rapid buildout of 5G network infrastructure is a
national security imperative. Future military preparedness will rely on
it, and fostering technologically competitive U.S. companies of all
sizes depends on it. Moreover, as the pandemic has made clear, strong
digital infrastructure bolsters our resilience to systemic shocks,
allowing Americans to access telehealth, education, and other services
they need in times of crisis. 5G networks will be the connective tissue
between all advanced mobile systems, and particularly in conjunction
with advances in AI and computing power, will enable profound new
technological capabilities directly in user devices. China has treated
this as a strategic priority and invested heavily in a Gigabit
nationwide mobile network, which it will soon achieve. In the United
States, however, 5G network development has proceeded slowly--only
delivering incremental increases in data speeds and coverage. We should
act now and decisively to improve the U.S. position. I have three
ideas.
First, we should reinvest spectrum auction proceeds into network
infrastructure. I suggest we examine ways to recycle the $81 billion in
revenue from the Federal Communication Commission's (FCC) Auction 107
of ``C-band'' spectrum, and any future auctions, into funding
designated for network infrastructure, with an allocation mechanism
designed to promote rapid and equitable buildout by the private sector.
Second, we should explore spectrum sharing and other auction
alternatives. For example, DOD has invited public input into how it
could share spectrum it controls with industry. I have suggested a
model wherein DOD retains control of the spectrum but allows industry
to share it in exchange for industry building the required
infrastructure quickly, and at its own cost. To be clear, this is not
``nationalized 5G,'' as some critics have claimed. This would be a
privately built, operated, and maintained network that prioritizes DOD
use. In any case, I believe DOD should be applauded for examining
innovative solutions to this urgent problem. Third, we should modify
auction terms. For any future auctions, particularly in the C-band
spectrum that is ideally suited for 5G, the FCC should impose strict
buildout requirements for auction winners that ensure that the
necessary network infrastructure gets built quickly and equitably. We
can't just wait for 6G or 7G to arrive. Competitive advantage
surrendered now is likely lost forever. I see this as an untenable
national security risk.
---------------------------------------------------------------------------
\3\ This approach could build on successful models such as the
COVID-19 High Performance Computing Consortium, (https://covid19-hpc-
consortium.org/) and NSF's CloudBank, (https://www.cloudbank.org/).
---------------------------------------------------------------------------
hardware vulnerabilities: microelectronics
After decades leading the microelectronics industry, the United
States is now almost entirely reliant on foreign sources for production
of the cutting-edge semiconductors that power all of the AI algorithms
that are critical for defense systems and everything else. The
dependency on semiconductor imports, particularly from Taiwan, creates
a strategic vulnerability from adverse foreign government action,
natural disaster, and other events that can disrupt the supply chains
for electronics--as we have seen in the auto industry recently.
Although American universities and firms remain global leaders in the
key areas of semiconductor R&D and chip design, the semiconductor
industry is now highly globalized and competitive. Taiwan Semiconductor
Manufacturing Corporation (TSMC) leads the world in semiconductor
contract manufacturing and Samsung in South Korea is also producing
state-of-the-art logic chips. Simultaneously, in a bid to catch up and
achieve chip self-sufficiency, China is pursuing unprecedented state-
funded efforts to forge a world-leading semiconductor industry by 2030.
If a potential adversary bests the United States in semiconductors, it
could gain the upper hand in every domain of warfare.
The United States should commit to a strategy to stay at least two
generations ahead of China in state-of-the-art microelectronics and
commit the funding and incentives to maintain multiple sources of
cutting-edge microelectronics fabrication in the United States. I would
recommend:
(1) the Executive Branch should finalize and implement a national
microelectronics leadership strategy; (2) Congress should offer a 40
percent refundable tax credit and grants for domestic fabrication
investments by firms from the United States and its allies; and (3)
Congress should appropriate an additional $12 billion over the next
five years for microelectronics research, development, and
infrastructure in key areas such as advanced packaging. These
investments should help accelerate the transition of ideas from
university prototypes to commercial-scale production domestically.
Together, these efforts will enable the U.S. Government, private
sector, and academia to rise to the challenge of rebuilding U.S.
semiconductor superiority. Focusing our efforts to develop domestic
microelectronics fabrication facilities will reduce dependence on
imports, preserve leadership in technological innovation, support job
creation, improve national security and balance of trade, and enhance
the technological superiority and readiness of the military--an
important consumer of advanced microelectronics.
implications of the emerging technology competition for defense
Emerging technologies are creating new whole-of-society threats.
This is not just, or even primarily, a traditional battlefield
challenge in the near term. AI-enabled capabilities will be tools of
first resort in a new era of conflict. State and non-state actors
determined to challenge the United States, but avoid direct military
confrontation, will use AI to amplify existing tools and develop new
ones. Adversaries are exploiting our digital openness through AI-
accelerated information operations and cyber attacks. ``Ad-Tech'' will
become ``NatSec-Tech'' as adversaries recognize what advertising and
technology firms have recognized for years--that machine learning is a
powerful tool for harvesting and analyzing data. Using espionage and
publicly available data, adversaries will gather information and use AI
to identify vulnerabilities in individuals, society, and critical
infrastructure.
Looking more narrowly at military issues, key technology areas have
important and wide-ranging defense applications. Fundamentally, the
sources of battlefield advantage will shift from traditional factors
like force size and levels of armaments, to factors like superior data
collection and assimilation, connectivity, computing power, algorithms,
and system security.
The advantages to be gained are well understood by our competitors.
Russia has plans to automate a substantial portion of its military
systems. China's military has embraced ``intelligentized war''--
investing, for example, in swarming drones to contest United States
naval supremacy. \4\ China is testing and training AI algorithms in
military games designed around real-world scenarios. The recent use by
Azerbaijan of drones and loitering munitions to defeat air-defense
systems and mechanized forces in Nagorno-Karabakh is a harbinger of the
kind the future American forces will soon face.
---------------------------------------------------------------------------
\4\ See Testimony of Elsa Kania before the United States-China
Economic and Security Review Commission, Hearing on Technology, Trade,
and Military-Civil Fusion (June 7, 2019), https://www.uscc.gov/sites/
default/files/June%207%20Hearing--Panel%201--Elsa%20Kania--Chinese
%20Military%20Innovation%20in%20Artificial%20Intelligence--0.pdf; Elsa
Kania, ``AI Weapons'' in China's Military Innovation, Brookings at 1
(April 20, 2020), https://www. brookings.edu/wp-content/uploads/2020/
04/FP--20200427--ai--weapons--kania--v2.pdf.
---------------------------------------------------------------------------
Defending against AI-capable adversaries without employing AI is an
invitation to disaster. AI will compress decision time frames from
minutes to seconds, expand the scale of attacks, and demand responses
that will tax the limits of human cognition. Human operators will not
be able to defend against AI-enabled cyber or disinformation attacks,
drone swarms, or missile attacks without the assistance of AI-enabled
machines. The best human operator cannot defend against multiple
machines making thousands of maneuvers per second potentially moving at
hypersonic speeds and orchestrated by AI across domains. Humans cannot
be everywhere at once, but software can.
The Pentagon is developing many operational concepts to fight these
future wars. But I am concerned that at the Department's current pace
of technology integration, the military will not be capable of carrying
them out in time. To fight as the military intends to fight in 2030 or
2035, the Department needs to get the fundamentals in place well before
then.
the commercial model
DOD needs to revise how it builds things. Silicon Valley has shown
a way to do this: form smart teams, drive hard deliverables, and move
quickly. The government does not allow any of that: procurement is
separate from design and design feedback, software is an afterthought,
and the big systems are siloed so they can't be integrated together. We
should build missiles the way we now build cars: use a design studio to
develop and simulate in software. Return to the skunkworks model of
fast iteration. The long design cycles are killing our competitiveness.
Fast iterative design and product cycles are the key to
competitiveness. DOD should target military systems that can be
accelerated by a new design studio and digital twinning approach and
change procurement rules to allow for it. At the very least DOD should
pick a few programs and agree collectively to run them very
differently.
getting the fundamentals right at the pentagon
I recognize I cannot wave a magic wand over the Pentagon, so below
are some important concrete things DOD should do now at a bare minimum.
Again, the NSCAI has detailed recommendations that I endorse for
getting the technical backbone right. These focus mainly on AI but most
have broad applicability for new technology integration and development
in DOD.
1. integrate existing digital technologies now
The Pentagon's byzantine processes can sometimes obscure a basic
point. Much of the new technology the military needs is already
available on the commercial market. Buy more of it. Doing so would
create market incentives to produce more and more useful defense
technologies. The Department should:
Prioritize existing technologies that can augment
intelligence functions--especially applications of AI. There are
significant opportunities to better leverage commercially available
technologies to improve situational awareness and indications and
warnings. Automation and human-machine teaming can enhance the
effectiveness of a range of ISR platforms and improve the full cycle of
intelligence collection and analysis.
Network DOD's digital innovation initiatives to scale
impact. A number of the Department's innovation organizations have
delivered results. \5\ But they are uncoordinated and under-resourced.
DOD signaling of technology priorities is ad hoc and is not supported
by a track record of significant DOD investments in digital technology
with non-traditional vendors. As a result, national security AI
applications attract less private-market investment. The Department
should harmonize its innovation initiatives to carry out a coordinated
strategy for commercial technology solutions. The Under Secretary of
Defense for Research and Engineering should direct this effort.
---------------------------------------------------------------------------
\5\ DOD innovation initiatives include various entities across the
military services and the Office of the Secretary of Defense that are
focused on bridging the gap with the commercial sector, especially with
start-ups and non-traditional vendors. These include the Defense
Innovation Unit, AFWERX, NavalX, and the Army Applications Laboratory,
among others.
---------------------------------------------------------------------------
Establish AI delivery teams at each Combatant Command. AI
delivery teams should be embedded at each Combatant Command and should
be capable of supporting the full lifecycle of AI development and
fielding--including data science, engineering, testing, and production.
Teams should include forward-deployable components to act as the local
interface with operational units.
2. improve the department's digital infrastructure
DOD took a promising first step in 2020 with the issuance of a Data
Strategy. \6\ However, the Department lacks the modern digital
ecosystem, collaborative tools and environments, and broad on-demand
access to shared AI resources it needs to integrate AI across the
organization. The Secretary of Defense should direct the establishment
of a DOD-wide digital ecosystem. The Secretary should require that all
new joint and service programs adhere to the design of this ecosystem,
and that, wherever possible, existing programs become interoperable
with it by 2025. This technical foundation should: 1) provide access to
leading cloud technologies and services for scalable computing; 2)
enable the sharing of data, software, and capabilities through well-
documented and hardened application programming interfaces with proper
access controls; and 3) give all DOD developers and scientists access
to the tools and resources they need to drive new AI capabilities.
---------------------------------------------------------------------------
\6\ See Executive Summary: DOD Data Strategy, U.S. Department of
Defense (Sept. 30, 2020), https://media.defense.gov/2020/Oct/08/
2002514180/-1/-1/0/DOD-DATA-STRATEGY.PDF.
---------------------------------------------------------------------------
At the same time, the Department should define a joint warfighting
network architecture by the end of this year. The goal should be to
create a secure, open-standards systems network that supports the
integration of AI applications at operational levels and across
domains. \7\ It should be accessible by all of the military services
and encompass several elements, including command and control networks;
data transport, storage, and secure processing; and weapon system
integration.
---------------------------------------------------------------------------
\7\ The network envisioned is well-aligned with ongoing DOD efforts
to embrace standards-driven interoperability, system adaptability, and
data-sharing. See Memorandum for Service Acquisition Executives and
Program Officers, U.S. Department of Defense (Jan. 7, 2019), https://
www.dsp.dla.mil/Portals/26/Documents/PolicyAndGuidance/Memo-Modular--
Open--Systems----Approach.pdf.
---------------------------------------------------------------------------
3. reform leadership structures
Leadership is the critical variable. Driving innovation requires
organizational change, not just technical capacity. Senior civilian and
military officials should set clear priorities and direction, empower
subordinates, and accept higher uncertainty and risk in pursuing new
technologies. Specifically, DOD should:
Establish a high-level Steering Committee on Emerging
Technology, tri-chaired by the Deputy Secretary of Defense, the Vice
Chairman of the Joint Chiefs of Staff, and the Principal Deputy
Director of National Intelligence. \8\
---------------------------------------------------------------------------
\8\ Section 236 of the Fiscal Year 2021 National Defense
Authorization Act allows the Secretary of Defense to establish a
steering committee on emerging technology and national security
threats. However, the structure described in Sec. 236 does not include
leadership from the Intelligence Community, which is critical to
ensuring a coordinated approach between DOD and the IC. See Pub. L.
116-283, William M. (Mac) Thornberry National Defense Authorization Act
for Fiscal Year 2021, 134 Stat. 3388 (2021).
---------------------------------------------------------------------------
Ensure the JAIC Director remains a three-star general or
flag officer with significant operational experience who reports
directly to the Secretary or Deputy Secretary. \9\
---------------------------------------------------------------------------
\9\ The Senate confirmed Lt. Gen. Michael Groen to lead the JAIC in
September 2020. NSCAI's has recommended that the three-star requirement
be statutorily mandated.
---------------------------------------------------------------------------
Appoint the Under Secretary of Defense for Research and
Engineering as the co-chair and chief science advisor to the Joint
Requirements Oversight Council.
4. build new talent pipelines
There is no conceivable program, pilot, internship or pathway for
tech talent that will close the DOD talent deficit, and the same
problem exists across all national security agencies. I cannot stress
enough the need for a radical rethinking of talent pipelines. The NSCAI
has exactly the right idea. This is not a time to add a few new
positions in national security departments and agencies for Silicon
Valley technologists and call it a day. We need to build entirely new
talent pipelines from scratch. We should establish a new Digital
Service Academy and civilian National Reserve to grow tech talent with
the same seriousness of purpose that we grow military officers. The
digital age demands a digital corps. Just as importantly, the United
States needs to win the international talent competition by improving
STEM education and our highly skilled immigration system.
Technology experts need better ways to spend a career in government
focused on their fields. Current talent management practices often put
experts in positions that are unrelated to their areas of expertise.
Many leave the government or military as a result. DOD should create
civilian and military career fields in software development, data
science, and AI. My philanthropy, Schmidt Futures, is sponsoring a
small pilot, called the Center for Digital Talent, that aspires to open
new recruiting pathways for technologists into the Department, but much
more work needs to be done.
Senior leader education is also very important. Leaders who do not
understand new technologies are less likely to pursue programs that
will add value. They will not be able to incorporate new technologies
into operational concepts or organizational processes. DOD should
create emerging technology critical billets and an emerging technology
certification process that is analogous to the current joint
qualification process.
DOD also needs to integrate computational thinking and AI basics
into junior leader training. NCOs and junior officers need a baseline
level of knowledge to responsibly field new capabilities. DOD needs to
integrate digital skills and computational thinking into pre-
commissioning requirements, initial officer training, and NCO
education. I recommend focusing on problem curation, data collection
and management, the AI lifecycle, probabilistic reasoning and data
visualization, and data-informed decision making.
5. invest more in s&t and align investments with strategy
The Department should commit to spending at least 3.4 percent of
its budget on science and technology, with a focus on emerging and
disruptive technologies. \10\ This would be a significant increase from
the current level of 2.3 percent, and would follow longstanding
recommendations by the Defense Science Board and others, which are
echoed in the forthcoming NSCAI report. For AI in particular, the
Department should increase R&D spending from around $1.5 billion to at
least $8 billion by 2025. \11\
---------------------------------------------------------------------------
\10\ This would encompass DOD budget activities 1 through 3, which
can help produce the advancements that will drive the next generation
of capabilities.
\11\ This should encompass investments in pushing the boundary of
AI technology towards new capabilities, and developing AI-enabled
elements to build into existing systems and platforms. The AI
Commission has identified a number of critical areas to be supported:
human-AI teaming; advanced scene understanding; intelligent edge
devices, computing, and networking; robust and resilient AI; AI test
and evaluation, verification and validation; integrated AI; modeling
and simulation for decision support; autonomous AI systems; advances
toward more general artificial intelligence.
---------------------------------------------------------------------------
To align investments with strategy, DOD should produce a Technology
Annex in the next National Defense Strategy document. This annex would
prioritize technology investments and development in relation to the
military capabilities needed to carry out future operational concepts.
And it would clearly signal which technologies are Department
priorities.
6. reform dod's outdated budget process
I've stated before that the DOD's problem is not innovation, but
innovation adoption. Its outdated, industrial-age budgeting process
creates a valley of death for new technology, allowing basic research
funding and also procurement of weapons systems, but preventing the
flexible investment needed in prototypes, concepts, and experimentation
of new concepts and technologies like AI.
Although we have had 50 years of acquisition reform, we have not
meaningfully changed the PPBE (Planning, Programming, Budget and
Execution) process developed in the 1960s. Congress and the Defense
Department need to work together to immediately authorize and fund
pilots, and set the stage for more sweeping reform. \12\
---------------------------------------------------------------------------
\12\ I am encouraged that the Fiscal Year 2021 NDAA included
support for the Department's Budget Activity 8 pilot program, which
seeks to overcome the barrier that DOD spending categories pose to the
development and sustainment of digital technologies. Congress and DOD
could build on this pilot to establish needed flexibility more broadly
by creating a single source of funding that could support the full
lifecycle of development, delivery, and continuous update for AI and
other digital technologies.
---------------------------------------------------------------------------
7. ensure responsible development, testing, and use of ai-enabled and
autonomous systems
I see a consensus emerging on how to use AI responsibly for
defense. The DIB produced a set of AI ethics principles. The AI
Commission followed with more granular, operational-level guidance.
These efforts have been well received by DOD leadership.
If an AI-powered system does not work as designed with
predictability and guided by clear principles, then operators will not
use it, the military services will not embrace it, and the American
people will not support it. Rushing to integrate AI would be
counterproductive if it caused service members to lose confidence in
its benefits. All military systems require rigorous testing,
safeguards, and an understanding of how they might operate differently
in the real world than in a testbed. AI-enabled autonomous weapon
systems could be more precise, and as a result, reduce civilian
casualties. But they also raise important ethical questions about the
role of human judgment in employing lethal force. If improperly
designed or used, they could also increase the risk of military
escalation.
An entirely new approach to testing, evaluation, validation and
verification (TEVV) will be needed. DOD should tailor and develop TEVV
policies and capabilities to meet the changes needed for AI as its AI-
enabled systems grow in number, scope, and complexity. This should
include establishing a TEVV framework and culture that integrates
continuous testing; making TEVV tools and capabilities more readily
available across DOD; updating or creating live, virtual, and
constructive test ranges for AI-enabled systems; and restructuring the
processes that underlie requirements for system design, development,
and testing.
conclusion
It has been a great privilege to have worked at the leading edge of
the American technology industry for over 30 years. That work began,
for me, with grants from the Federal Government.
My graduate work in computer science in the 1970s and 1980s was
funded in part by the National Science Foundation and the Defense
Advanced Research Projects Agency. These and other investments fueled a
renaissance of technology that made America and its technology sector
the envy of the world and our military the most capable fighting force
in history.
But right now, the United States is not playing to win. It is the
Chinese who are competing to become the world's leading innovators.
Never before in my lifetime have I been more worried that we will soon
be displaced by a rival or more aware of what second place means for
our economy, our security, and the future of our nation.
A bold, bipartisan initiative can extend our country's technology
advantage but only if we act now. Success matters for more than our
companies' bottom lines and our military's battlefield edge. Because
our technology and that of our closest allies and partners embodies our
values, advancing individual liberty and strengthening free societies
are also on the line. I leave you with the urgent message that for the
American model to win, the American Government must lead. To that end,
I urge Congress again to adopt all of our AI Commission
recommendations, which provide a clear blueprint to win a technology
competition that is centered around AI.
Chairman Reed. Thank you very much. Mr. Smith, please.
STATEMENT OF MR. BRAD L. SMITH, PRESIDENT, MICROSOFT
CORPORATION
Mr. Smith. Well, thank you, Mr. Chairman and members of the
committee, and let me join Eric in saying thank you for having
this hearing and giving us the opportunity to share our ideas
with you.
Let me build on what Eric has said, because I think he
covered a lot of things extremely well. Clearly technology is
changing every aspect of society, including the nation's
national security needs. It starts with the cloud and the edge
and it goes to 5G and AI and a future based on quantum
computing. And I think the first question for all of us is
really to ask, how should we, as a nation, think about what
this means for the defense of the country in the future?
I think the answer is really with a combination of
confidence and concern. I think there are many reasons to be
confident, and, Mr. Chairman, you referred to some of them. We
have the world's best research universities. We have an
enormously creative and dynamic commercial technology sector.
We have a military that both quantitatively and qualitatively,
on a person-by-person basis, is the best in the world. And
perhaps most importantly, we stand for democratic principles
and values that most of the world, quite rightly, wants to
follow.
That is a formidable combination, and yet I do believe
there are causes for concern, really two. Eric covered the
first well. We are competing with a formidable competitor.
China is investing, and it is investing heavily in every area
of technology we are here to talk about this morning.
But I think there is a second dimension as well. Over time,
technology either favors offensive weaponry in attacks or
defensive protection against attacks. And if you think about
American history, geography has always been our friend. We
could look not to one large ocean but two to keep our
adversaries at a distance. But the truth is the internet has
changed all of that. It has made everybody each other's next-
door neighbor.
And I think we should draw a lesson, even from the events
of the last week. Think about what happened when the electrical
grid went down in Texas. Think about the danger to American
civilians if there is a disruption of the water supply. And
then think about a future where a nation need not send missiles
or planes but can simply send code to do its fighting for it.
This is changing the threat landscape, and unfortunately favors
offensive attacks against a very broad defensive horizon that
must be secured.
So what do we do? Well, Eric has already touched on a
number of important ideas. I would mention four. Number one, we
need to strengthen the nation's digital infrastructure and
digital defenses, and that touches every part of the public
sector and every part of the private sector as well. Number
two, we need to think about and decide how we can harness these
advances in technology to equip our warfighters in the nation's
military it can move faster and continue to be at the
technological edge.
Certainly at Microsoft we have had the opportunity to do
that in recent years. We have had the opportunity to work with
the Department of the Army on the Integrated Visual
Augmentation System goggles that provide not only night vision
and thermal vision but lots of other data as well. And we have
seen the Army benefit from the procurement reforms that this
committee has advanced, and believe it, it changes everything,
in my view, about how we can innovate faster.
Number three, we need to think not just about military
applications but the health of our technology base as a whole--
the education of our people, the investment in higher education
and research, our immigration system, and how we advance the
areas of technology where we risk most falling behind.
And finally, we need to work more closely with our allies
than ever before, and we need to lead with moral authority and
not the strength of technology alone. We need to remember every
day that there will never be perhaps another day when we will
be competing with an adversary that has a smaller population
than ours. But we do, in fact, have a set of human rights
protections and democratic values that can pull the world
together. And when we succeed in doing that, both to harness
the power of our technology and to build an alliance of
partners and friends, I think we put this country on the course
that it needs, that should give us all more confidence than
concern.
Thank you very much.
[The prepared statement of Mr. Smith follows:]
Prepared Statement by Brad Smith
Chairman Reed, Ranking Member Inhofe, and Members of the Committee,
thank you for the opportunity to offer some perspectives on issues that
are vital to U.S. national security.
Digital technology plays an increasingly critical role in the
defense of the nation. Emerging technologies are redefining the way we
secure the peace, maintain our defense, and when necessary, fight wars.
Innovations in cloud and edge services, artificial intelligence,
and 5G are already having a direct and practical impact on the nation's
defense. As the decade progresses, we should look to the potential
importance of quantum computing as well. These technologies will
redefine the requirements for military operations at mission speed,
based on their ability to harness massive amounts of data and
computational power. They will also be interconnected: future
computational capabilities will be defined by an ability to accelerate
applications across the cloud, using AI and advanced silicon. They will
reshape the security needs for the nation's critical infrastructure and
affect training requirements for our military personnel. In short, new
technology will have a pervasive impact on our national security.
Yet one would be hard-pressed to say that the country currently has
a comprehensive strategy to harness these technologies for the
country's defense. A more cohesive approach is needed.
This strategy needs to be grounded in a clear-eyed assessment of
where digital technology is going and the nature of global competition
in technology markets. Speed matters. The United States must move more
quickly to advance broad-based technology innovation and pursue new
approaches to use, secure, and adapt commercial advances for military
applications. This requires a holistic approach to government-sponsored
basic research, commercial technology development, and investments in
new military uses. It will require an even closer partnership between
the government and the tech sector.
An essential starting point is to ask: What are we trying to
accomplish? To be sure, the protection of American lives and the peace
and prosperity of our country are the primary considerations. But so is
the country's unique role in providing global leadership. When we think
about the role of technology in the context of the country's defense
and national power, our ability to lead the world and to establish and
defend the most important connective tissue of the international
order--in areas such as finance, cybersecurity, healthcare, and
transportation--marks one of the deepest roots of American power and
security.
For the last 70 years, the United States has provided what we might
think of as the global public operating system in every essential area
of life. The next 70 years will witness this not as a metaphor, but as
real software power. Any successful national security strategy
therefore must also find ways for us to continue to offer the best
options for nations around the world as they transition every part of
their national lives to a digital age. As in the past, there is no
substitute for technology the world can trust, based on the United
States' commitment to human rights and democratic values.
Based on this vision, the country should pursue a digital defense
strategy with seven objectives:
1. Focus on where digital technology is going and where advantage
will lie.
2. Strengthen the nation's technology leadership by investing in
talent and research.
3. Enhance American competitiveness and security by modernizing
technology-related trade and investment policy.
4. Accelerate the adaptation of commercial digital technology for
defense applications.
5. Continue to strengthen the defense of the nation's digital
infrastructure.
6. Pursue a strong and renewed commitment to technology
collaboration with our allies.
7. Lead with moral authority and not the strength of technology
alone.
All this is described in greater detail below.
1. focus on where digital technology is going and where advantage will
lie.
Almost all the digital technology we rely on today was made
possible because of Gordon Moore's simple rule: processing speeds of
silicon chips double every two years. For more than a half century,
this principle has defined the explosive advancement of hardware,
software, and connectivity. While Moore's Law is reaching the physical
limits of fabricated chips, computing will continue to advance at a
rapid rate. Today's focus on algorithms, software, new materials,
integration technologies, and even subatomic research will redefine
computing. And while the computer revolution took root on American
soil, it is now a worldwide endeavor with global powers, including
China, competing and sometimes leading the race.
The rise of the cloud and the transition to distributed intelligence at
the cloud and the edge.
Cloud services have become the lifeblood of most modern
enterprises. They make large amounts of computational power and storage
available without capital investments in hardware by the end user. This
is reshaping military technology in the same way it is impacting every
other field. DOD has embraced these trends through projects like the
Joint Enterprise Defense Initiative (``JEDI'').
JEDI lays the foundation for DOD to embrace a full array of
transformative technologies. For the first time, DOD will be able to
fully leverage billions of dollars of annual private sector investment
in cloud security, reliability, infrastructure, and governance. It will
replace investments in single-purpose systems that are out of date by
the time they come into service, using instead a modern compute
environment that evolves with changing technology. Investment in hybrid
solutions will further enable these core capabilities to extend from
the data center to the field with new devices that enable data insights
and analysis in rugged environments with poor or no connection to the
network.
As the cloud extends its reach beyond data centers to what has been
coined the ``intelligent edge,'' cloud computing is becoming
geographically distributed through an ever-expanding Internet of Things
(IoT). Whether in a home, vehicle, or factory, the edge is considered
one of the last bastions of Moore's Law as embedded sensors and devices
become more efficient and less expensive. By 2030, 50 billion IoT
devices will reside on the edge of the world's computing network. Just
two years from now, in 2023, International Data Corporation (IDC)
projects that more than 50 percent of new enterprise IT infrastructure
will be at the intelligent edge rather than corporate data centers, up
from less than 10 percent in 2020. By 2024, the number of applications
deployed in the cloud and at the edge will increase 800 percent. \1\
---------------------------------------------------------------------------
\1\ Macy Bayern, ``IDC: Top 10 Worldwide IT Predictions for 2020,''
TechRepublic, October 29, 2019, https://www.techrepublic.com/article/
idc-top-10-worldwide-it-predictions-for-2020/.
---------------------------------------------------------------------------
This means the future of computing for everything, including
military applications, is about the combination of computing power in
the cloud and at the edge, with robust connectivity between them. As
Microsoft CEO Satya Nadella has noted, the acceleration of this type of
``tech intensity'' is essential for any institution to thrive going
forward. One has to be both world class at adopting the latest digital
technology and building its own proprietary digital technology. This is
going to be true for our defense institutions as well. \2\
---------------------------------------------------------------------------
\2\ Satya Nadella, ``The Necessity of Tech Intensity in Today's
Digital World,'' LinkedIn, January 18, 2019, https://www.linkedin.com/
pulse/necessity-tech-intensity-todays-digital-world- satya-nadella/;
see also Dr. Tianyi (TJ) Jiang, ``#TechIntensity Explained: 4 Ways It
Shifts Business Strategy Forever,'' AvePoint Blog, October 10, 2018,
https://www.avepoint.com/blog/office-365/tech-intensity-explained/
#:8:text=%E2%80%9CTech%20intensity%E2%80%9D%20is%20a%20
phrase%20coined%20by%20Microsoft,its%20ability%20to%20build%20its%20own%
20digital%20 capability.
---------------------------------------------------------------------------
As the world's intelligent edge explodes, so will the amount of
data gathered by the tiny sensors and devices located where the digital
and physical worlds intersect. Paired with Artificial Intelligence (AI)
and its use of Machine Learning (ML), edge devices will have the power
to see, listen, reason, and predict real-world developments around
them. Perhaps more importantly, new intelligent edge applications will
be able to interact with their physical environments to perform
increasingly complex tasks with increasing degrees of autonomy. And as
intelligent devices at the edge proliferate, so too will the surface
area for cyberattacks as the vulnerabilities of these soft access
points are exploited.
This distributed paradigm will bridge the physical and digital
worlds by enabling previously difficult or impossible scenarios, like
digital twins and rich real-time analytics to support our military on
the most remote battlefields. Microsoft and the U.S. Army have already
moved forward on this digital frontier by working together on the
Integrated Visual Augmentation System (IVAS), based on the HoloLens 2
augmented reality (AR) system. For example, before warfighters seek to
rescue hostages in a building, they can plan their mission based on a
digital twin of the building and train for the operation using the
rapid construction of a physical mock-up of it. The same technology
enables warfighters to execute the operation with real-time visual data
that integrates everything from the building's digital layout to local
thermal images to facial recognition of the hostages and the
identification of friendly forces.
Connectivity--from broadband to 5G, Low-Earth Orbit Satellites, and
beyond.
As the computing canvas stretches from the edge to the cloud,
reliable connectivity will become essential to provide the bandwidth
and speed needed to maximize smart and connected devices. Fifth-
generation, or 5G, networks will deliver data flows 10 to 100 times
faster than 4G and support many more devices. They will offer the
precision and speed needed to realize the power of edge computing with
immersive, real-time, and intelligent experiences, much like
electricity powers the world today.
Countries that rapidly deploy 5G stand to gain in revenue, job
creation, and leadership in technology innovation. As we have seen with
other technology transformations, software will play an important role
in advancing 5G to deliver new solutions that increase speed, reduce
costs, and boost security. With 5G more so than previous generations of
wireless technology, software--from signal processing to radio area
networks to complex traffic management--is at least as critical as
spectrum and radio frequency infrastructure.
There is a significant opportunity for both traditional leaders and
new players across the industry to innovate, collaborate, and create
new markets, serving the world's networking and edge computing needs
and the coming software ecosystem that will depend on these
technologies. As with previous technology ecosystems, global standards
and interoperability in our networking and computing infrastructure
across the edge and cloud will be critical to unlocking the full
creativity and productivity of the scientists, engineers,
entrepreneurs, and innovators who will help shape our future.
As nations look to overhaul their broadband infrastructure,
governments are rightly focusing on the cyber risks associated with
5G's supply chain integrity where they currently rely exclusively on a
handful of foreign suppliers. While some nations are breaking this
dependency by adopting modularized software-defined systems, some are
concerned that these systems create a broader and multidimensional
vulnerability. 5G's inherently modular nature and use of software-
defined networking, however, also create opportunities to increase
security and resiliency. This can foster a more diverse supplier
ecosystem and enable the application of leading-edge security
techniques and technologies, such as AI and containerization to
identify, isolate, contain, and protect against malicious attacks on
the network.
But 5G is not the only connectivity technology that is advancing.
Existing solutions like fiber, satellites, Wi-Fi, and short-range
technologies continue to progress. For example, we can leverage
satellite broadband to connect modular data centers to bring high-
intensity, secure cloud computing to some of the most challenging
environments, where critical prerequisites like power, connectivity,
and building infrastructure are unreliable. And well before we reach
the year 2030, we'll be discussing 6G and how to extend the networks'
global reach through thousands of Low-Earth Orbiting Satellites.
Software combined with the explosion of data and infused with AI.
For most of the 179 years since Lady Ada Lovelace wrote the first
program for a computing device--Charles Babbage's Analytical Engine--
software programming has required a skilled individual to translate a
human's understanding of a problem to a program that instructs a
machine how to solve it. AI, particularly with the stunning progress
computer scientists and engineers have made in ML over the past two
decades, has allowed us to think about harnessing computers in a
fundamentally different way.
ML systems learn from data without being programmed. They can
reason about complex phenomena in both the digital and physical world,
understand these phenomena, and make predictions or draw inferences
that can support human decision making or be employed in automated
ways. Using ML techniques, we have built AI systems that can both see
and understand what they are seeing. We have built speech recognition
systems that can hear and understand what is being said. We have built
systems that can seamlessly and in real time translate between spoken
human languages. We have even built AI systems that have achieved
superhuman performance on tasks we once thought were high watermarks of
human cognition, like beating the best human players in the world at
games like Chess and Go.
The power of machine learning systems is growing rapidly, both in
terms of improved performance on existing ML tasks (like speech
recognition, computer vision and machine translation), and perhaps more
interestingly, on the rapid expansion of new tasks that ML systems can
undertake.
AI and machine learning workloads that run side by side with more
traditional, hand-coded software will continue to grow at an
exponential rate, driven by developers utilizing new AI algorithms and
customers' ambitions to incorporate AI into new tasks. According to
IDC, by 2024, more than 50 percent of user interface interactions will
use AI-enabled computer vision, speech, natural language processing
(NLP), and augmented and virtual reality (AR/VR). \3\ And by 2025, at
least 90 percent of new enterprise applications will embed artificial
intelligence.
---------------------------------------------------------------------------
\3\ International Data Corporation, ``IDC FutureScape Outlines the
Impact `Digital Supremacy' Will Have on Enterprise Transformation and
the IT Industry,'' October 29, 2019, https://www.idc.com/
getdoc.jsp?containerId=prUS45613519.
---------------------------------------------------------------------------
Recent ML breakthroughs, particularly the family of methods jointly
referred to as deep learning, have allowed ML systems to approach or
exceed human capabilities on a wide range of tasks. These breakthroughs
enable us to teach AI systems to accomplish a very broad range of
cognitive tasks by training on unlabeled data, such as Wikipedia texts
and YouTube videos. Given the extremely large volumes of unlabeled data
available on the internet, as well as data that can be produced in
simulation environments and will be coming in growing volumes from
sensors on the intelligent edge, we increasingly are bound more by the
amount of computing power than the amount of data that we can bring to
bear to train ever-larger models. Researchers anticipate that this
trend will continue to yield results even as models grow to be 100 to
1,000 times larger than they are today.
The need to train models this large has unleashed a new race to
create ``AI supercomputers,'' with a primary competitive race unfolding
between Google, based in part on its acquisition of DeepMind, and Open
AI, which works with a substantial investment from and in partnership
with Microsoft. As this race has progressed, Google, Open AI, and
Microsoft have achieved new landmark results in natural language
processing with AI models that now have hundreds of billions of
machine-learning parameters. It has also led to additional
breakthroughs in computer vision, speech recognition, content
understanding and recommendations, and other areas of machine learning.
The implications for defense applications are expanding rapidly.
For example, Microsoft is leveraging commercial AI technology to
accelerate innovation for DOD through the creation of computer-
generated, three-dimensional models of objects and environments. Until
recently, Pentagon planning often was constrained by the availability
of imagery from the theater of operations. Leveraging technology
developed by our Xbox team, we combine gaming and rendering technology
developed for consumer markets over the last 20 years to build lifelike
models depicting objects in any environment, at any time of day, in any
weather condition, and from any angle or perspective. (This technology
is also being used in civilian scenarios to train Unmanned Aerial
Vehicles to recognize the state of crops to enhance productivity for
farming.) DOD can use these models to train personnel and plan
operations.
Semiconductor Chips--from faster processing speeds to a quantum future.
While the first generation of AI supercomputers are being built
with today's most powerful semiconductor chips and networks, the
building blocks for these systems were not originally designed to
support AI at scale. The next generation of AI supercomputers will
require a surge in innovation in silicon, computer architecture,
memory, and networking technology. Tomorrow's AI supercomputers will
need to be orders of magnitude more powerful than the most powerful
machines in existence today to meet the nearly unbounded demand for
compute from modern AI programs.
As our need for compute continues to expand, the physical
limitations of silicon are becoming apparent, spurring research and
development on materials with enhanced capabilities to support new
forms of computation, including quantum computing. Classical computers
powered by silicon think in terms of binary bits of ones and zeroes.
Quantum computers, by contrast, harness modern physics and the quantum
mechanical behavior of nature to perform a computation using quantum
bits--or qubits--the quantum version of a classic binary bit that
represents multiple values simultaneously.
The promise of quantum computers lies in their ability to solve
problems requiring ``big compute''--challenges in cryptanalysis,
chemistry, and materials science--in months, weeks, or days, where
current and even the next generations of silicon-based chips and
networks would still take billions of years. Once scaled up, quantum
computing could lead to rapid advances across society and industry,
including identifying an efficient catalyst to reduce carbon dioxide in
the atmosphere and materials that could enable lossless power
transmission or better battery technologies.
Unlocking the full potential of quantum computing will require more
than simply building quantum computers, however. Quantum applications
will require advanced classical computers working in conjunction with
quantum computers. These applications on an industrial scale will
require advances in semiconductor chips, cloud infrastructure, network
connectivity, and more.
It is important to both national security and the American economy
to secure a domestic quantum future. The National Quantum Initiative
Act signed in 2018 was a critical first step. It bolstered the nation's
leadership by investing in quantum research and development by
government, industry, and academia. Industrial-scale quantum computing
will require even more, including a physical infrastructure to support
the quantum supply chain that encompasses manufacturing, materials
development, system-level validation and verification, and nanoscale
fabrication.
Looking to the future, Congress should consider funding a quantum
equivalent of Operation Warp Speed. The U.S. Government could seek to
combine federal resources with private sector capital and expertise.
Federal funding could come in the form of milestone-based pre-payments
for access to the capabilities that firms are developing, direct
funding for scalable quantum solutions, and other means of accelerating
and de-risking quantum efforts. Congress should also consider ways to
increase cooperation and knowledge sharing between government quantum
researchers and their private sector counterparts.
The conceptual threads that tie American technology together.
The foregoing areas reflect an enormous range of scientific and
technological advances. Yet two conceptual threads run throughout all
these critical fields. First, advances and adoption of technology at a
global level require more than world-class technology itself. They also
turn on the ability to persuade other governments and international
markets to adopt standards and endorse technology protocols that
reflect American inventions. The United States has excelled in these
fields through decades of international collaboration and outreach. It
will need to continue to do so for decades into the future.
Second, and perhaps more important, all these innovative
technologies require and run on trust. As digital technology becomes an
ever more ubiquitous part of our lives, it has increasingly profound
impacts on our privacy, safety, security, and other fundamental
freedoms. This too has deep implications for American leadership and
values. Global technology competition is not only about the latest
technical invention. It is also about products that reflect values the
world can trust.
2. strengthen the nation's technology leadership by investing in
talent and research.
National policy for digital defense technology also needs to be
grounded in a clear-eyed assessment of the state of global technology
markets and the nature of global technology competition. There are two
factors that deserve special attention.
American digital defense technology increasingly starts with the
development of commercial technology and then moves to military
and intelligence adaptations, rather than the other way around.
Since the 1800s, military technology has fallen into two
categories. The first is illustrated by the jeep, a classic example of
commercial technology that the military adapted for use in World War
II. Henry Ford debuted the Model T in 1908 as the world's most
practical and inexpensive automobile. Ford and other American
automakers improved on this design for decades. In 1940, the U.S. Army
recognized that the approaching war would require a new and inexpensive
four-wheel drive motor vehicle. It turned to the nation's
manufacturers, who adapted off-the-shelf automotive parts and designed
the first prototype in just two days.
The other category is technology that is invented first for
military use and subsequently adapted for commercial applications. A
good example is the jet aircraft. America's first jet plane was the
Bell P-59 Airacomet, also created during World War II. It was designed
in secret and its invention wasn't shared with the public until 1943,
after it had completed 100 flights. It would take 15 additional years
before the jet engine would be attached to civilian aircraft and
transform the world of commercial aviation.
The Cold War and the race to the moon were won principally by
technology developed first for the government and later put to
commercial use. But today the sequence often is reversed. Developing
digital defense technology is often more like designing jeeps than
inventing jets. This phenomenon, in turn, creates a need for American
leadership in two areas--world-leading technology research and
development capability in both the governmental and private sectors and
the ability to quickly adapt civilian technology for military use.
The country must continue to refresh its capacity for digital
innovation by investing in talent and research.
The United States is the world's technology leader today because of
decades of investment in education and research. When the nation
confronted the Sputnik launch by the Soviet Union in 1957, President
Eisenhower and a bipartisan Congress recognized that sustained national
progress required not just federal investment in a new generation of
rockets, but in stronger math and science education for a new
generation of people. \4\ Just 11 months after Sputnik's launch,
Eisenhower signed the National Defense Education Act (NDEA) into law,
saying ``this emergency program stems from national need, and its
fruits will bear directly on national security.'' \5\
---------------------------------------------------------------------------
\4\ Wayne Urban, More than Science and Sputnik: The National
Defense Education Act of 1958 (Tuscaloosa: University of Alabama Press,
2010); Yanek Mieczowski, Eisenhower's Sputnik Moment: The Race for
Space and World Prestige (Ithaca, NY: Cornell University Press, 2013).
\5\ Dana Adrienne Ponte, The First Line of Defense: Higher
Education in Wartime and the Development of National Defense Education,
1939-1959 (Seattle: University of Washington Unpublished PhD
Dissertation, 2016), 89.
---------------------------------------------------------------------------
Federal investment in education and basic research created a
powerful infrastructure for innovation, but like our roads and bridges,
that infrastructure is showing its age. Last month, the National
Security Commission on Artificial Intelligence (NSCAI), chaired by my
co-panelist Eric Schmidt, said in its Draft Final Report that ``the
time is right for a second NDEA, one that mirrors the first
legislation, but with important distinctions.'' \6\ This frames the
issue well and rightly sets a high bar for the bold ambition the
country needs to refresh its innovation infrastructure. A new federal
initiative should include the following elements, among others:
---------------------------------------------------------------------------
\6\ National Security Commission on Artificial Intelligence, Draft
Final Report, January, 2021, 82, https://www.nscai.gov/wp-content/
uploads/2021/01/NSCAI-Draft-Final-Report-1.19.21.pdf.
---------------------------------------------------------------------------
Expand support for STEM education. Today, less than a
third of American high schools offers an advanced placement course in
computer science. \7\ The number of young people taking such a course
in 2020 was lower than for eleven other subjects. One challenge is the
high cost of training teachers to teach computer science. Philanthropic
groups such as code.org and tech companies such as Microsoft, Google,
and Amazon have all launched important initiatives to help address this
need, but more federal leadership and funding is needed, especially to
support teacher training.
---------------------------------------------------------------------------
\7\ College Board, ``AP Program Participation and Performance Data
2020,'' https://secure-media.collegeboard.org/digitalServices/pdf/
research/2020/Program-Summary-Report-2020.pdf.
Invest in post-secondary education for critical
disciplines. Federal support under the NDEA targeted disciplines such
as math and science (and especially physics) that Congress believed
would be critical to winning the space race. A similar effort is needed
today, and it should start by cataloguing the fields where there is a
current or expected shortage of skilled personnel in the United States.
This should address the need for a compute-savvy workforce skilled in
key areas like AI, quantum, and cybersecurity. Like the NDEA itself,
this effort should include a focus on career and technical education,
leveraging the nation's community colleges and vocational schools as
---------------------------------------------------------------------------
well as four-year colleges and graduate degree programs.
Modernize immigration laws to address technology needs.
The country's last major immigration overhaul took place in 1986, when
Ronald Reagan was President and Tip O'Neil was Speaker of the House. It
was closer in time to Sputnik's launch in 1957 than the technology
challenges of 2021. The NSCAI's Draft Final Report captures well the
types of immigration changes that are needed to ensure the United
States attracts the best and brightest talent needed to advance
technology's frontier. These include broadening the visa category for
extraordinary talent, enabling better job portability for highly
skilled visa holders, and enacting measures to clear the current green
card backlog and provide a more stable path to green cards in the
future. \8\ In addition, we should not forget that the Nation's
Dreamers include a substantial number of extraordinarily talented
individuals with advanced technology skills, something we witness every
day among DACA registrants who are Microsoft employees.
---------------------------------------------------------------------------
\8\ National Security Commission on Artificial Intelligence, Draft
Final Report, 82-86.
Increase federal support for basic research related to
critical technologies. The United States retains an unmatched
capability for basic research through the country's research
universities. Yet United States Government spending on research and
development and our share of global spending have dramatically
declined, \9\ and within the next few years China is expected to
surpass us. \10\ As in the past, the country needs to bolster our
research capability for the next generation of technology needs,
including AI, quantum computing, and other critical technologies. Here
too, the NSCAI gets it right in its Draft Final Report, recommending an
increase in AI R&D at compounding levels, doubling annually to reach
$32 billion per year by fiscal year 2026. \11\
---------------------------------------------------------------------------
\9\ Congressional Research Service, The Global Research and
Development Landscape and Implications for the Department of Defense,
November 8, 2018, https://crsreports.congress.gov/product/pdf/R/R45403.
\10\ Paul Scharre and Ainikki Rikkonen, Defense Technology
Strategy, Center for a New American Security, November 2020, CNAS--
Defense Technology Strategy (s3.us-east-1.amazonaws .com)
\11\ NSCAI Draft Final Report, 90.
Support DOD efforts to recruit tech talent and develop
digital skills among DOD personnel. Finally, the decade ahead will
require that every American employer, including the nation's military,
do more to invest in digital skills for its own personnel. While the
country's employers increased their investments in digital skilling
between 1980 and 2000, these investments have fallen and then stagnated
since the year 2000. \12\ Part of what is needed for the future will
involve heightened DOD recruiting of tech talent. Virtually every job,
including virtually every position in our military, will require more
digital skills a decade from now that it does today. And conversely, as
servicemembers exit the military, we need to support them to move into
technology-enabled roles so their national security experience can help
drive private sector applications and innovation. A successful example
of a public-private partnership in this area is the Microsoft Software
and Services Academy (MSSA). \13\ It has enabled thousands of service
members, veterans, and spouses to secure technology jobs with more than
600 employers across the country.
---------------------------------------------------------------------------
\12\ Brad Smith, ``Microsoft Launches Initiative To Help 25 Million
People Worldwide Acquire the Digital Skills Needed in a COVID-19
Economy,'' Microsoft on the Issues (blog), June 30, 2020, https://
blogs.microsoft.com/blog/2020/06/30/microsoft-launches-initiative-to-
help-25-million-people-worldwide-acquire-the-digital-skills-needed-in-
a-covid-19-economy/.
\13\ Microsoft Corporation, ``Microsoft Software and Services
Academy'', https://military. microsoft.com/programs/microsoft-software-
systems-academy/.
3. enhance american competitiveness and security by modernizing
technology-related trade and investment policy.
The United States has been a global leader in digital technology
since the field's inception, but this leadership will be more
challenging to maintain in the decade ahead. While this conversation
often begins by comparing the tech sectors in the United States and
China, it is helpful to start by identifying the factors that influence
this competition more broadly.
American success in the development of commercial technology typically
requires success on a broad international scale.
This is true for three reasons. First, digital technology often
involves high fixed costs and low marginal costs. The fixed costs are
for engineering involved in software development and capital costs such
as the construction of chip fabrication or data center facilities. To
charge low prices and gain market share, companies must spread these
high fixed costs across a large customer base that can only come from
growth in foreign markets.
Two other factors are at work as well. Most technology markets have
strong network effects, which enable strong returns to scale once a
company has established market leadership. And finally, services that
are dependent on large quantities of data for product improvement,
including through ML, are likely to gain an additional advantage by
being the first to reach a market leading position. All this explains
why LinkedIn founder and Microsoft board member Reid Hoffman talks
about the critical need in tech markets for blitzscaling, meaning a
``lightning-fast path'' to develop market leadership on a global scale.
\14\
---------------------------------------------------------------------------
\14\ Reid Hoffman and Chris Yeh, Blitzscaling: The Lightning-Fast
Path to Building Massively Valuable Businesses (New York: Currency,
2018).
---------------------------------------------------------------------------
This has implications for competition between the American and
Chinese commercial technology sectors. With a population of 1.4 billion
people, China is in a unique position to develop technology markets at
an unmatched domestic scale. The rapid growth of ByteDance's TikTok
service illustrates this well. As of last year, the company's service
inside China, named Douyin, had 600 million daily active users, while
its international TikTok counterpart had another 689 million monthly
active users, giving it almost 1.3 billion users worldwide. \15\ This
same phenomenon is at work for Chinese companies that are marketing
technological platforms to global consumers in areas such as
healthcare, finance, and education.
---------------------------------------------------------------------------
\15\ Brian Dean, ``TikTok Demographics Statistics: How Many People
Use TikTok in 2021?'', Backlinko, November 4, 2020, https://
backlinko.com/tiktok-users.
---------------------------------------------------------------------------
At the same time, American technology firms do not have full access
to China's domestic technology market. This makes it even more
important that American companies succeed quickly not only in the
United States, but in many other international markets as well.
The United States currently has a patchwork of technology-related trade
and investment laws rather than a holistic, cohesive, and
strategic regulatory approach.
Last summer Microsoft had not just a front row seat but a direct
participatory role in some aspects of the TikTok review. One thing we
came to appreciate is the difficulty for government officials and
private sector participants alike when making decisions about specific
technologies in the absence of a clearer overall legal framework to
guide technology-related trade and investment activities. The United
States' current patchwork of laws in these areas not only lacks
strategic coherence but also reduces predictability for everyone it
affects.
On the export front, Congress in 2018 enacted the Export Control
Reform Act (ECRA), the most sweeping piece of export control
legislation since the 1970s. While this legislation directed the
Commerce Department's Bureau of Industry and Security (BIS) to adopt
new regulations, the process for doing this--still ongoing--is creating
substantial uncertainty for the tech sector. This is a critical and
ongoing issue for almost every large technology company in the United
States, as firms seek to balance these compliance obligations with the
demands of a global market that wants more American products ever
faster--and where missing a single product cycle can make it very
difficult to catch up.
On the import front, United States policy has moved rapidly to
restrict technology investments and imports from China. This has its
roots in the Committee on Foreign Investment in the United States
(CFIUS), established in the 1970s. Congress expanded CFIUS's
jurisdiction in 2018 through the Foreign Investment Risk Review
Modernization Act (FIRRMA), which authorizes the Committee's review of
non-controlling foreign investments. The National Defense Authorization
Act for Fiscal Year 2019 requires federal contractors to ban certain
telecommunications technologies from their supply chains. The last
Administration also relied on the International Emergency Economic
Powers Act (IEEPA) to broadly authorize Commerce's review of technology
transactions and ban certain mobile applications.
In recent years, the government has relied on this complex set of
laws to address several technology-related concerns. Some of these
efforts have focused on specific companies and the technologies they
provide. Others have involved broad categories of information and
communications technologies. For example, the State Department in
recent years encouraged other countries to adopt more restrictive
policies in these areas through its ``Clean Network'' initiative.
It is worth recognizing that China's policies in this area reflect
a similar desire to manage technology trade. China has long had a
restrictive legal regime to manage technology imports and investments.
This combines the filtering of foreign content with an array of
domestic licensing requirements, joint venture obligations, and
informal government signaling regarding the purchase of foreign
technology. Last August, the Chinese Government adopted new rules to
control technology exports as well. These measures substantially
broaden controlled categories, now including social media algorithms
and other new categories. These changes were followed by a new export
control law that went into effect in December, representing China's
most significant effort to date to implement a comprehensive ``dual-
use'' export control regime.
As we look to the decade ahead, it is apparent that both the United
States and China will want to scrutinize and restrict trade in dual use
technologies. And with an increasing focus on digital sovereignty, the
European Union and some member states are moving in a similar
direction.
Given the stakes and uncertainty, the urge to err on the side of
caution by adopting ever more restrictive policies in this space is
understandable. But that approach could weaken national security by
undermining American technology leadership. We need a balanced and
coherent framework that will protect national security without
isolating the United States. And as we consider issues related to China
in particular, we should develop an approach to technology-related
trade and investment that permits cooperation when it is clearly in the
interest of American technology leadership. As modern as China may be
today, the country still depends on American technology and standards.
To pull away from that position and accelerate China's adoption of its
own, competing approaches risks jeopardizing American leadership in
critical areas.
The country needs to modernize its technology trade and investment
policies.
The Commerce Department should identify the commercial
technology exports it wants to control and adopt a modern, calibrated
approach to control them. A high priority for the Commerce Department
should be the adoption of new regulations on ``emerging and
foundational technologies'' under ECRA. As many companies across the
tech sector noted last year, applying a traditional, restrictive export
control approach based solely on a product's performance criteria not
only risks limiting societally beneficial uses, but could hinder the
development of new technologies by depriving companies of the scale
necessary to compete internationally. Overly restrictive export
controls also risk cutting off access to the best talent--not just from
the country targeted for control, but also from allies and other like-
minded nations.
A new and more calibrated approach is needed. Microsoft and Open AI
proposed one in comments submitted to Commerce in November. Under this
proposal, the Commerce Department would set policies that determine who
can access sensitive technologies and for what purpose. \16\ This would
allow for protection against problematic users and uses in a more
targeted, effective, and dynamic way--not just at the point of initial
access but continuously in a deployed environment. These policies would
then be implemented and enforced within the protected technology
itself, as well as by hardening the infrastructure around it to prevent
circumvention.
---------------------------------------------------------------------------
\16\ Sarah O'Hare O'Neal, ``Microsoft and OpenAI Partner to Propose
Digital Transformation of Export Controls,'' Microsoft on the Issues
(blog), November 10, 2020, https://blogs. microsoft.com/on-the-issues/
2020/11/10/openai-partnership-digital-export-controls/.
---------------------------------------------------------------------------
New technologies make this approach feasible. For example, software
features built into sensitive technologies can enable real-time
controls against prohibited uses and users. These features would
include identity verification systems and information flow controls.
``Tagging'' can be used to ensure the same controls apply to
derivatives of these sensitive technologies.
Similarly, ``roots of trust'' built into sensitive hardware
technologies can require authorization to send code or data through the
equipment. More robust hardware identity verification through secure
co-processors akin to those used to secure payment in mobile phones or
to prohibit in-game cheating in game consoles can further protect
hardware against unauthorized access and uses.
Technology may not eliminate the need for restrictive export
controls in every particularly sensitive scenario. But in many areas,
more targeted, technology-enabled solutions could help strike an
optimal balance between security and the need for the American
technology sector to remain globally competitive.
The government should ensure there exists an independent
supply chain for both existing and certain anticipated critical
technologies. To address this challenge, at least two key questions
await urgent answers.
First, the country must decide what technologies should be provided
exclusively from domestic sources or from allied nations. The key
criteria likely should focus both on the sensitivity of the technology
and the danger of supply disruption in the event of international
tensions. For example, the United States currently cannot source
critical 5G technologies in a cost-effective way either domestically or
from allies. It is impossible to imagine our potential adversaries
being comfortable relying exclusively on American suppliers for these
same technologies. The United States shouldn't think about these issues
any differently.
Second, the United States must decide how to achieve supply chain
independence in the selected technologies in a strategic, effective,
and cost-efficient way. Some key tenets should guide this work. First,
the government should take stock of market trends and build upon them,
providing public financial support only where it is needed and in a
manner that will accelerate sustainable development by the market
itself. Second, the government should use the full range of its policy
tools to accelerate essential market trends, including its procurement
practices and the broad role in the economy of agencies such as the
DOD. And finally, the government should ensure there is reciprocal
trade access to the American market for suppliers from NATO and other
allied democratic countries, based on common terms for American access
to these other markets.
The United States should modernize its broader technology
import and foreign investment policies. This goes beyond the question
of where the country wants to have an independent source of supply.
Instead, it asks the government to decide where the presence of certain
foreign technologies and investment poses a threat to the country's
national security.
The challenge of managing technology imports is more daunting than
for exports, in part because there has been no legislation in recent
years akin to ECRA. While IEEPA is a powerful policy tool, it was
developed in a different era and for different circumstances from those
that exist today. On the investment front, Congress recently updated
CFIUS. But the United States still lacks a coherent framework governing
the related issues of technology imports and foreign investment in U.S.
technology companies. There are several critical questions that require
an answer.
First, the government must decide which technologies are so
sensitive that imports or foreign ownership need to be controlled. It
should then adopt consistent policies to manage both imports of and
foreign investments in these technologies. The technology horizon will
continue to evolve rapidly, and the government therefore will need
criteria that stands the test of time. In part this should include
digital infrastructure that would be susceptible to penetration or
disruption in times of war.
Second, once these sensitive technologies have been identified, the
government must decide how it wants to control them. While one approach
would be to bar sensitive technologies or investment from certain
countries entirely, this is not always the best or the only feasible
approach. For example, Microsoft has long operated by creating
transparency centers that enable appropriate inspection of source code
for a product like Windows. Similarly, we developed last year and
shared with United States officials what we regarded as a sophisticated
and effective technology model to manage consumer services from China
by addressing five key objectives--security, privacy, authenticity,
digital safety, and transparency.
It is likely that global trade in key sectors increasingly will
rely on these types of technology-enabled solutions. The United States
should become an early adopter so that it can lead and shape the
development of these solutions internationally.
Finally, just as the government must determine where to restrict
technology trade, it should also identify certain areas where it is
safe for technology to move freely across borders. The good news is
that many technologies are not sensitive from a national or economic
security perspective. Even more important, in an era of open-source
code and broad-based basic research, human knowledge advances daily
based on global collaboration. The United States should aspire to lead
the world in advancing the frontiers of scientific understanding and
spreading appreciation of humanitarian values. We need government
policies that protect the country's national security without cutting
ourselves off from the global conversations that will shape humanity's
future.
4. accelerate the adaptation of commercial digital technology for
defense applications.
The biggest competitive challenge the United States confronts in
competition with China is not in technology research and development.
Instead, it is the advantage China has over the United States in faster
deployment and adoption of new technologies, particularly in AI. There
are multiple reasons for this, including China's centralized government
direction, and to some degree, broad adoption of technologies in ways
that Americans rightly find objectionable. But one unmistakable result
is the need to encourage faster and broader deployment and adoption of
emerging technologies in the United States in a manner consistent with
democratic principles and American values. This includes more rapid
adoption of emerging digital technologies by DOD, most importantly to
ensure American military supremacy but also to help accelerate
technology adoption more broadly.
Our national security will be best served through a three-pronged
effort by the government to utilize digital technology. First, the
government should use commercially available technology when it is
sufficient for the task and as the foundation for additional
development when more work is needed. This will both accelerate speed
and reduce costs. Second, the government should add security layers to
commercial technology when required, such as by protecting secret and
top-secret workloads and military operations. Third, the government
should adapt commercial products and development methods for military
uses and applications, including through additional product development
of the sort illustrated by the IVAS.
All these efforts should be guided by three goals, among others--
speed, cost, and innovation. As discussed further below, there is an
opportunity to build upon recent procurement reforms with additional,
practical steps that advance these goals. As much as anything else, we
need to build a foundation for rapid and creative co-development
efforts that breaks down barriers between engineers in the private
sector and the warfighters whose missions depend on effectively using
the world's most advanced technology.
This Committee has pursued critical and impactful work in recent
decades to reform DOD procurement. Much of this work has focused
rightfully on the shift from the hardware-centric weapons systems of
the post-World War II and Cold War eras to the digitally enhanced
technologies of the 21st Century. Despite this progress, there remain
important inefficiencies that collectively impede DOD's ability to
rapidly adopt digital technology. From incentives that reward process
over speed to protests that undervalue the urgency of deploying the
newest innovations, the Pentagon is still not where it needs to be. DOD
should adopt approaches that will:
Incentivize and train the acquisition workforce. In the
private sector, we see risk-taking, failure, and iteration as a natural
part of the innovation process. The DOD acquisition workforce, on the
other hand, is more heavily incentivized to be risk averse. This should
change. DOD should recruit, train, and retain the tech talent needed to
develop, test, integrate, and deploy new technologies. It should reward
this professional procurement corps for agility, speed, smart risk-
taking, and accountability.
This Committee was instrumental in passing the Other Transaction
and Middle Tier Acquisitions authority and procedures designed to
dramatically speed up the adaptation of commercial technologies for
defense use. The IVAS program is a case study in this innovation.
Nonetheless, we see added opportunity for the Pentagon to take
full advantage of the tools this Committee has given it. There are
still days when parts of the Pentagon find comfort in the rigidity of
the Federal Acquisition Regulation (FAR) over the speed and flexibility
of these newer tools. Even when new procurement channels are used, the
process is sometimes managed in ways that resemble the more restrictive
and slower processes the new channels were designed to replace. The
future mission-critical capabilities needed for battlefield superiority
will require that those responsible for requirements, acquisition, and
technology deployment all work together faster, more closely, and
seamlessly--and in conjunction with private sector innovators.
Create an Innovation Infrastructure. A recent report by
the Center for a New American Security \17\ found that the Pentagon
lacks a robust digital infrastructure to support modern warfighting.
Building this infrastructure will require additional investments in
cloud computing, data labeling and storage, and the human capital
needed to fully utilize and manage these tools.
---------------------------------------------------------------------------
\17\ Michele Flournoy and Gabrielle Chefitz, Sharpening the U.S.
Military's Edge: Critical Steps for the Next Administration, Center for
a New American Security, July 13, 2020, https://www.cnas.org/
publications/commentary/sharpening-the-u-s-militarys-edge-critical-
steps-for-the-next-administration.
---------------------------------------------------------------------------
The 2018 DOD cloud strategy \18\ noted that ``the DOD
information environment is made up of multiple disjointed and stove-
piped systems distributed across modern and legacy infrastructure
around the globe.'' A more unified general purpose cloud environment is
a key prerequisite for breaking down these barriers and speeding up the
adoption and development of transformative technologies.
---------------------------------------------------------------------------
\18\ Department of Defense, DOD Cloud Strategy, December 2018,
https://media.defense.gov/2019/Feb/04/2002085866/-1/-1/1/DOD-CLOUD-
STRATEGY.PDF.
Review and reform the government contract protest
process. The government contract protest process needs to be reformed
to strike a better balance between fairness and open competition, on
the one hand, and the urgency of innovation on the other. The existing,
outdated process often leads to uncertainty, extended delays, and
protracted litigation, hindering the speed of innovation and often
maintaining the status quo. When considering acquisition reforms,
Congress should look at ways to modernize, streamline, and accelerate
protest actions. These should include time limits, not only on the
filing of protests but on case resolution and corrective actions.
Concluding bid protests more quickly will help provide our warfighters
the technology they need when they need it.
5. continue to strengthen the defense of the nation's digital
infrastructure.
For two centuries, technology has been changing the nature of what
is needed to defend a nation. The first two years of the 1940s
illustrate this well. In early 1940, the tank rendered worthless two
decades of French investment in the Maginot Line, as it was suddenly
possible for an Army to go around it. And in late 1941, the United
States learned that advances in naval aviation meant that battleships
could no longer defend Pearl Harbor. If not defended against
effectively, foreign cyberweapons pose a similar threat of comparable
severity in our current day.
Nature's recent impact in Texas demonstrates the potential
devastation that would result if a foreign adversary used cyberweapons
to take down a nation's electrical grid. Yet it has been apparent since
2014 that Russian agencies have been targeting the United States
electrical grid. \19\ And in 2017 the citizens of Ukraine experienced
an even broader cyberattack that was launched by disrupting the
software supply chain, in that case through malware implanted in an
update for local accounting software. As one author has noted, ``in the
cyber world, what happens in Kiev almost never stays in Kiev.'' \20\
The recent malware attack on SolarWinds demonstrates the truth of at
least part of this proposition.
---------------------------------------------------------------------------
\19\ David Sanger, The Perfect Weapon: War, Sabotage, and Fear in
the Cyber Age (New York: Crown, 2018), 163-68.
\20\ Ibid., 163.
---------------------------------------------------------------------------
These issues also reach our democratic infrastructure, connecting
national needs that are as old as our Republic with the most modern
technology of the 21st Century. As George Washington recognized in his
Farewell Address, democratic societies depend on a unique combination
of free expression and social cohesion that must be protected from
foreign interference. \21\ Yet recent years have seen Russian successes
in turning American social media into a Weapon of Mass Confusion,
illustrated by the 2016 success of the Internet Research Agency in St.
Petersburg in organizing a synchronized protest and counterprotest in
Houston. \22\ The nation's digital defenses today must include stronger
measures to protect against disinformation campaigns, the misuse of
personal information, and the voting process itself.
---------------------------------------------------------------------------
\21\ George Washington, ``Washington's Farewell Address of 1796,''
Avalon Project, Lillian Goldman Law Library, Yale Law School, https://
avalon.law.yale.edu/18th--century/washing.asp.
\22\ Brad Smith and Carol Ann Browne, Tools and Weapons: The
Promise and the Peril of the Digital Age (New York: Penguin Press,
2019), 96.
---------------------------------------------------------------------------
The DOD and other parts of the U.S. Government have made rapid
progress in addressing many of these issues in recent years, but there
remain several new and additional priorities that should be addressed,
including:
Strengthen supply chain security for the private and
public sectors for both software and hardware. The public sector at all
levels of government should strengthen the protection of their
software, including through secure development practices, better
software maintenance and vulnerability management, and integrity
controls that apply throughout the software development, testing, and
delivery processes. The implementation of this year's National Defense
Authorization Act provides an opportunity to develop new software
acquisition security requirements that may be appropriate across
federal agencies.
Broaden use of cybersecurity best practices, including
through improved cyber hygiene and a commitment to IT modernization.
The public sector in the United States needs to continue to modernize
its technology base, in part through cloud migration that can better
ensure ongoing state-of-the-art software code and improved threat
detection. This should be coupled with the broader adoption of strong
security practices such as the establishment of a Zero Trust
environment, assessments of the security of cloud providers, and the
re-orientation of risk management activities to complement third party
services and security automation.
Develop a national strategy to strengthen the sharing of
threat intelligence across the entire security community, including
through a clear, consistent disclosure obligation on the private
sector. Much as radar advances proved indispensable in helping to
defend against air attacks in World War II, modern threat intelligence
can help defend against cyberattacks today. But only if threat
intelligence is shared quickly and effectively. There is a critical
need to improve the sharing of threat intelligence across the Federal
Government, with key American allies, and in an appropriate but
collaborative way with tech companies that often are cybersecurity
first responders. This also requires consideration of new measures to
ensure that attacks on private enterprises are reported in an
appropriate way to a federal agency, consistent with the protection of
personal privacy.
6. pursue a strong and renewed commitment to technology collaboration
with our allies.
The United States cannot secure its digital defenses by acting
alone. One of the country's greatest strategic advantages is its global
network of allies and partners. In part this is because of the global
nature of technology innovation and markets. Microsoft's quantum
computing efforts illustrate this well, with cutting-edge labs in
Indiana, California, and Washington, as well as in Denmark, the
Netherlands, and Australia.
Moreover, as noted above, one of the key drivers of successful
development and deployment of technology is scale. The larger the
potential market for U.S. technologies, the larger the pool of private
and human capital that will be dedicated to the research and
development efforts needed to maintain America's competitive edge.
Scale plays a major role in AI development, in particular. AI runs on
data. That means that China, with a population of 1.4 billion, has a
comparative advantage when it comes to mustering sheer quantities of
personal data. But the combined populations of the United States, our
NATO and Five Eyes allies, Japan, and Korea, total over 1.1 billion. If
Mexico, India, and Brazil are added, the combined population of this
potential coalition of democracies would be close to 2.9 billion.
The United States should work with its global network of alliances
and partners to:
Invest in and build coalitions with like-minded partners
to develop, adapt, and deploy new technologies. In part this should
include selected basic research programs, like those discussed above,
to bring together NATO members, the Five Eyes, and other democratic
allies. It should also include efforts to align our technology trade
policies and laws, as discussed above, with those of our allies.
Address privacy issues that undermine trust across the
Atlantic. There is a pressing need to address a short list of high
priority privacy concerns, starting with improvements to the U.S.-E.U.
Privacy Shield. These efforts should build a foundation for more
durable global solutions to address issues around government access to
data and should include international agreements under the CLOUD Act
with the European Union and other American allies.
Pursue an ambitious digital and technology trade agenda.
The United States should build on the landmark digital trade rules in
USMCA by upgrading other free trade agreements to include rules on data
localization, cross-border data flows, and forced disclosure of
proprietary source code and algorithms. At the same time, the country
should continue to push for high-standard outcomes in the ongoing WTO
digital trade negotiations. It should also explore the possibility of
an even more ambitious plurilateral digital trade pact with like-minded
countries and seek cooperation on standards for a range of emerging
technologies.
Advance strong norms for global cybersecurity protection.
The United States should embrace international standards such as the
Paris Call for Trust and Security in Cyberspace, already endorsed by
more than 75 governments and more than 1,000 other signatories. It
should similarly advance norms for the cybersecurity protection of
software supply chains in the United Nations and elsewhere.
7. lead with moral authority and not the strength of technology alone.
Finally, while the United States will remain a preeminent economic
power for the foreseeable future, we must recognize that the nation no
longer retains one of the strategic advantages it enjoyed for much of
the 20th century, namely an economy that was orders of magnitude larger
than its principal rivals. In addition, the country must grapple with
one of the biggest challenges confronting the nation's defense--the
need to preserve bipartisan and broad support for our national security
policy in an era defined by a polarized public and a divided world.
Yet the country retains an enormous strength and strategic
advantage. When the United States stands firmly for its historic
democratic principles and the protection of human rights, it speaks and
acts with a moral authority that none of its adversaries can match.
There are few institutions that reflect and embody this strong ethical
tradition better than the United States military. It is an asset that
provides a strong cornerstone for future national and global
leadership, and the country needs to nurture and build on it further.
As Microsoft and so many other tech companies experience every day,
a new generation of Americans asks not only what will make their
country strong but their society great. It is the type of question that
should inspire us to be bold in our ambitions. As we've found, it is
critical to talk with our employees about the American military's
strong ethical traditions. When we do this and share our commitment as
a company to provide the U.S. military with the best technology we
create and simultaneously use our voice to advance ethics for AI,
almost uniformly our employees do not object. They applaud. Literally.
It is this type of appreciation that enables a company like Microsoft
to recruit top tech talent internally and externally for an ``all-
volunteer'' and ``all-star'' team for a project like IVAS. This type of
understanding also helps to strengthen America's technology leadership
through the active engagement and support of this country's technology
talent, as well as people who are not American by birth or citizenship.
Continued leadership in technology will require that we meet the
ongoing challenge to make sure American democratic principles and
values are an integral part of developing and deploying the next
generation of technology. This should include the following:
Continue to strengthen ethical practices and policies for
DOD's use of AI and other new technologies. DOD's adoption last year of
ethical principles to govern the use of AI not only represented a
critical step forward for the United States but also defined an ethical
role model for the world. Building on the recommendations of the
Defense Innovation Board, these principles sent a powerful message by
stating that military personnel ``will exercise appropriate levels of
judgment and care, while remaining responsible for the development,
deployment, and use of AI capabilities.'' \23\ The DOD principles also
addressed the importance of reliability, safety, transparency, and
bias. The Joint Artificial Intelligence Center (JAIC) is already taking
steps to implement these practices broadly. The NSCAI similarly has
offered additional and important ideas to implement ethical AI
principles throughout DOD and other agencies. The U.S. Government
should continue this work and discuss it broadly with the American
public.
---------------------------------------------------------------------------
\23\ C. Todd Lopez, ``DOD Adopts 5 Principles of Artificial
Intelligence Ethics,'' DOD News, February 15, 2020, https://
www.defense.gov/Explore/News/Article/Article/2094085/dod-adopts-5-
principles-of-artificial-intelligence-ethics.
DOD should encourage the adoption of similar ethical
principles and practices by its allies. The United States should
exercise its moral authority by encouraging NATO and other allied
nations to adopt similar ethical principles for their own militaries'
use of artificial intelligence. The AI Partnership for Defense (AI PfD)
announced last year between the United States and twelve allied nations
can serve as a forum for these discussions. The government similarly
should advance human rights norms and safeguards for new technologies,
including the use of facial recognition and government access to
---------------------------------------------------------------------------
personal information.
DOD should integrate environmental sustainability
concerns into its policies and practices. Finally, climate and energy
issues are having and will continue to have major consequences on our
national security. DOD has significant opportunities to substantially
enhance resilience, reduce carbon emissions, and catalyze innovation
though its own operations and supply chain. Many of these opportunities
are enabled by digital transformation. Cloud computing tools not only
can lead to significant operational energy and carbon efficiency gains,
but also provide key information for security landscape assessments in
countries around the world.
* * *
The challenges described above are formidable. But with concerted
effort, appropriate investment, and strong leadership from members of
this Committee and others, the United States can maintain its
competitive edge in technology and secure the nation's defenses. I look
forward to your questions and welcome the opportunity to discuss how
Microsoft and other technology companies can assist in these efforts.
Thank you.
Chairman Reed. Thank you very much. General Carlisle,
please.
STATEMENT OF GENERAL HERBERT J. CARLISLE, USAF (RET.),
PRESIDENT AND CHIEF EXECUTIVE OFFICER, NATIONAL DEFENSE
INDUSTRIAL ASSOCIATION
General Carlisle. Chairman Reed, distinguished members of
the committee, thank you for this opportunity to share my
experiences and industry perspective on emerging technologies
to ensure that our nation continues to be the preeminent force
in the 21st century. I would like to echo my colleagues'
comments of we really appreciate the opportunity to spend time
with you and give you our perspectives and help our nation move
forward in this area.
The last time I testified was during my final tour in
uniform, where I had the honor of leading Air Combat Command at
Langley Air Force Base. In that role, I was responsible for
organizing and training combat-ready forces. Before assuming
command of ACC, I was the commander of Pacific Air Forces,
responsible for all Air Force activities in about half of the
globe.
During my 40 years of service, I witnessed firsthand
numerous technological advances that focused on ensuring our
warfighters operate with the best, most innovative equipment to
ensure they are never in a fair fight. From my first flight in
a T-37, a long, long, long time ago, to my final flight in a F-
15, technological advances helped our forces go faster,
farther, and safer with greater lethality. My role at the
National Defense Industrial Association (NDIA) continues that
mission, to work with you and your esteemed colleagues, the
Pentagon, and the hundreds of thousands of members of industry
who strive to imagine and create the best and most advanced
equipment and capabilities to arm those young women and men
that serve our nation today.
We are almost a quarter of the way into the 21st century
and the character of war has changed somewhat. The threat to
our nation's defense is not necessarily countering state and
non-state actors in the domains, but it is looking at
cyberspace and actual outer space, and how we defend in those
areas. I think adversaries know they cannot outlast our
American industrial might today, but they are making gains in
changing the calculus every single day. Our competitors get
stronger, unconstrained, frankly, by fiscal year budgets, and
continuing resolutions are continuing to be a challenge. The
2018 National Defense Strategy identified 11 bipartisan
modernization priorities, including hypersonics,
microelectronics, and directed energy. We all agree that these
11 priorities are the emerging technologies priorities.
We know our peer competitors are investing in these areas
extensively, especially China. I have to say, I served in the
Pacific Theater throughout my career. Much of my 40 years was
in the Pacific, and as a squadron commander and in Pacific in
the early '90s, China was essentially a third-world nation. We
really did not consider them a legitimate threat at the time.
As Pacific Air Force (PACAF) commander in the 2010s, they were
not just a rising threat; they became, and are today, the
pacing threat. China has made particular inroads in hypersonics
by outspending us, outpacing us, and building on our work.
China's ambitious plans in space have led them to make
incredibly rapid advancements. They seek to build a
microelectronics capability within their nation. Even now, they
can very rapidly put state-of-the-art components into their
equipment, while U.S. military systems, in some cases, are two
generations behind. In some areas, like rare earths, we have
already fallen behind and are dependent on others. In other
areas, the question is no longer whether our adversaries will
close the gap, but whether we will catch up.
Where our competitors can, they have stolen our technology,
and where they cannot they have used predatory investments,
directed investments, and compulsory cooperation between
domestic and military in their countries. DOD needs to utilize
all the tools they have and adjust a risk-averse culture. Fewer
regulations, with more uniform enforcement, will ease the
burden on companies and the Department and speed up the
acquisition process. A workforce empowered and given authority
to make decisions provides the opportunity to unleash
innovative companies. This may lead to some failures in
programs and some long terms, but DOD can take a page from the
corporate world and learn from research and development (R&D)
failures.
We need to encourage and expand new and innovative
partnerships across government, industry, and academia to
exploit the pace of innovation and rapidly scale
transformational research and operational prototyping. We have
several mechanisms with which to do this and field products
quickly. We have small business innovation research (SBIR), we
have Defense Advanced Research Project Agency (DARPA), Defense
Innovation Unit (DIU), Space Development Agency, AFWERX,
SOFWERX, and many more. They demonstrate daily they can bring
nontraditional players into the defense industrial base in a
timely manner.
We need to be nimble and thoughtful, encouraging the
services to identify and support the transition of world class,
disruptive technologies.
Ladies and gentlemen, we truly appreciate congressional
support in helping DOD adopt an approach to accepting risk
intelligently--it is taxpayer dollars and we have to be smart--
taking a more collaborative approach across services to
identify and deploy game-changing technology that allows the
Department to maximize our limited resources. The men and women
in uniform sacrifice daily to protect our nation, our freedoms,
and our way of life. They deserve every protection that we can
afford them, and the equipment, capabilities, and training to
do the missions this nation asks them to do.
Thank you again for the opportunity to be here today. Thank
you, sir.
[The prepared statement of General Carlisle follows:]
Prepared Statement by General Herbert ``Hawk'' Carlisle
introduction
Chairman Reed, Senator Inhofe, and distinguished Members of the
Committee thank you for the opportunity to share my experiences and
industry perspective on emerging technologies so that we will ensure
our nation continues to be the preeminent force for the 21st Century.
The last time I testified was during my final tour in uniform,
where I had the honor of leading Air Combat Command (ACC). In that
role, I was responsible for organizing, training, and equipping combat-
ready forces for rapid deployment and employment while ensuring forces
were ready to meet the challenges of peacetime air sovereignty and
wartime defense. Before assuming command of ACC, I was the Commander of
Pacific Air Forces, responsible for Air Force activities spanning more
than half the globe.
During my 40 years of service, I witnessed firsthand numerous
technological advances that focused on ensuring our warfighters operate
with the best, most innovative equipment to ensure they are never in a
fair fight. From my first flight in a T-37 to my final flight in a F-
15, technological advances helped our forces go faster, farther, and
safer with greater lethality. My role at the National Defense
Industrial Association (NDIA) continues this mission--to work with you
and your esteemed colleagues, the Pentagon, and the hundreds of
thousands of members of industry who strive to imagine and create the
best and most advanced equipment and capabilities to arm our men and
women today, with an eye to what future engagements will require.
Almost a quarter of the way into the 21st Century, the character of
war has changed. The threat to our nation's defense is not necessarily
boots on the ground in far-off theaters; we're securing our networks
and countering state and non-state actors in the domains of cyberspace
and actual space. Adversaries know they cannot outlast American
industrial might today. But they are making gains in changing this
calculus. Our competitors get stronger every day--unconstrained by a
fiscal year budget and without the concerns of possible continuing
resolutions to inhibit their ability to innovate while placed in a
budgetary holding pattern. The 2018 National Defense Strategy
identified 11 bi-partisan modernization priorities, including
hypersonics, microelectronics, and directed energy, which we agree are
the right focus.
We know our peer competitors are investing in these areas as well,
especially China. I served in the Pacific Theater throughout my
uniformed service. As a squadron commander and in the 1990s, China was
essentially a third-world nation we did not consider a genuine threat;
as PACAF commander in the 2010's, they were not just a rising threat;
they became the pacing threat. China has made particular inroads in
hypersonics by out-spending us, out-pacing us, and building on our
work. China's ambitious plans in space have led them to make incredibly
rapid advancements. They are also investing heavily in AI and
biotechnology. They seek to build a domestic microelectronics
capability, but even now, they can put state-of-the-art components in
their systems, while US military systems are two generations behind. In
some areas, like rare earths, we have already fallen behind and are
dependent on others. In these areas, the question is no longer whether
our adversaries will close the gap, but whether we will catch up to our
competitors.
Where our competitors can, they've stolen our technology, and where
they can't, they've used predatory investments, massive directed
investments, and compulsory cooperation from domestic industry through
military-civil fusion. Combating these predatory economics requires a
whole of nation approach to both protect and promote American industry
to support our warfighters. From a defense industrial policy
perspective, this includes identifying ways to efficiently and
effectively deliver from research and development to acquisition, from
commercial as well as traditional defense firms, to pull forward and
not leave this technology to solely the commercial market or sitting on
the shelf. DOD needs to utilize the authorities they have and adjust a
risk-averse culture. Fewer regulations, with more uniform enforcement,
will ease the burden on companies and the agency and speed up the
acquisition process. A workforce empowered and given authority to make
decisions provides the opportunity to unleash innovative companies.
This may lead to some dead or wrong turns, but DOD can take a page from
the corporate world and learn from R&D failures. Strengthening the
transition of SBIR investments into programs of records is one such
method. On the promote side of the ledger, we need to make sure the
Department is a customer of choice for emerging technology providers.
This will require acquisition processes that operate at the speed of
relevance and budget stability so we can send a clear demand signal so
industry can effectively plan and commit resources.
We need to encourage and expand new and innovative partnerships
across government, industry, and academia to exploit the pace of
innovation and rapidly scale transformational research and operational
prototyping into robust and scalable capabilities that will enable
technological, and operational, superiority. We have several mechanisms
to develop and field products quickly and in an innovative manner;
SBIR, DARPA, DIU, the Space Development Agency, AFWERX, demonstrate
daily they can bring nontraditional players into the DOD in a timely
manner. We need to be nimble and thoughtful, encouraging the services
to identify and support the transition of world class, disruptive
technologies.
As part of the acquisition strategy, review prior SBIR projects and
assess opportunities to utilize SBIR authorities or projects. Leverage
the agile, time-saving authorities resident in the SBIR Phase III
contracting to get those technologies under contract and delivering to
the warfighter.
We appreciate Congressional support in helping DOD adopt an
approach to accepting risk intelligently. Taking a more collaborative
approach across services to identify and deploy game-changing
technology prevents duplicative efforts and allows the Department to
maximize limited resources. The men and women in uniform sacrifice
daily to protect our nation, our freedom, and our way of life. They
deserve every protection that we can afford them.
Chairman Reed. Thank you very much, General, and gentlemen,
thank you for your excellent testimony.
Before we begin questions, since we have some of our
colleagues that are attending remotely I want to let everyone
know how we will conduct the hearing. Since it is not possible
to know exactly when our colleagues who will be joining by the
computer arrive, we will not be following the standard early-
bird timing rule. Instead, we will handle the order of
questions by seniority, alternating sides until we have gone
through everyone. Once we reach the end, if there is anyone we
missed we will start back at the top of the list and continue
until everyone has had their turn. We will do the standard
five-minute rounds, and I ask my colleagues, particularly those
virtually attending, to keep an eye on the clock, which you
should see on your screens.
Finally, to allow for everyone to be heard, whether in the
room or on the computer, I would ask all colleagues to please
mute your microphone when not speaking. Thank you very much.
We were chatting before, and reminiscing about days gone
by, and one of the relics of those days gone by is the current
DOD budget process, the PPBE, Planning, Programming Budgeting
and Execution. It was a product of the McNamara, the Whiz Kids,
and I can assure you those Whiz Kids are not kids anymore. It
is 70 years.
So I will ask all the members, beginning with Dr. Schmidt,
do you think we need to modify this process in order to provide
the kind of organizational responsiveness, and are there any
other specific recommendations in terms of the current programs
and doctrines of DOD that you would suggest, Dr. Schmidt.
Dr. Schmidt. So there are a lot of problems with the
current procurement process, Mr. Chairman, and as a result,
every few years there is a redo of them, which just makes it
more complicated. There was a joke that the only way to
understand the procurement process was to have an AI system
explain it to everybody, I am sorry to say, but that is the
joke.
There are a number of problems with it. One has to do with
its design cycle. There is something called a POM, or a program
of record. There is a two-year planning cycle ahead of actually
approving anything. So if you want to do something new, you
have to plan it, and then it starts two years from the time you
get it, because that is when you get the money for it. Because
of the way the appropriators work, money that is not used in a
particular time is taken away unless it is on an identified
POM-based program.
This structure means that the people who should be making
the decisions, who, in my opinion, are the combatant commands
(COCOMs) and the heads of, you know, the Secretary of the Army
and Navy and so forth and so on, find that they do not really
have control over what is going on. They are responsible but
they do not have the ability to affect these things. The result
is the procurement systems are typically increasing. Every
generation is increasing two years in design cycle, and the
costs, of course, go up.
There are a number of mechanisms that you all have given
the DOD over the years which are special authorities of one
kind or another, and one of the questions that I do not
understand is why, if you give them the special authority, they
do not take it. So what I would suggest is that you give them
more authorities and you also ask them to try to figure out why
they are not taking advantage of them, because we are all in it
together to get faster design cycles.
To give you an idea, and I will finish, to give you an idea
of how strange the current design cycle is, in a normal
business you would have an idea, you would have the engineers
and the product people, you would have a chat, you would figure
out how much it costs. The CEO or product person would say,
``Let's do it.'' That is precisely not what happens in the
military. There is a requirements document, which is not
allowed to be communicated to the people who actually are going
to build it. There is no feedback between the people building
it and the actual requirement document. As a result, the
requirement document gets longer and longer and longer, and the
requirements cause the tradeoffs to get more and more
complicated, and you end up with a camel rather than a horse.
And that is the overall cycle, and that is why these
systems are so incredibly expensive. Changing that would save
money and it would make us much, much more effective.
Chairman Reed. Mr. Smith.
Mr. Smith. I would offer two ideas, one, building on what
Eric said. I think the more we can encourage the Pentagon to
use the authority that you have created for some of these
emerging technologies, the better off the nation will be. I am
not here to say that you buy an aircraft carrier the same way
you build software, but it is clear in the software space that
you can be agile. And what we have found, in a very, I think,
positive way, is when you can bring software developers and,
say, warfighters together, so they iterate in a training site,
and the warfighters identify a feature they need, and the
developers go off and create it over the next day, and then
they try it again, you can suddenly enable the military to move
forward at the speed of technology. So that is something worth
pursuing further.
And then second, I do think it is a good moment in time to
step back and look at our protest process. The protest process
today definitely does not move forward at the speed of
technology. And we all want to ensure fairness, and that
includes a fair right to be heard. But we could definitely
benefit from an accelerated timeline to do so.
Chairman Reed. Thank you. General Carlisle, please.
General Carlisle. Thank you, Mr. Chairman. I could not
agree more with my colleagues on the panel. The problem that I
faced when I was making decisions, I was a programmer in the
Air Force so sadly I know PPBE very, very well and very
painfully. And because of its two-year process there are so
many people that can stop it along the way. There are so many
levels that you go through.
So how you flatten that that is referenced as a suggestion
we have an office in the Air Force called the RCO, the Rapid
Capabilities Office. And the head of the RCO has authorized
money to work on programs and goes directly to the Secretary of
the Air Force, with nobody in between. And that ability to
flatten that and get it done more rapidly is really a
suggestion moving forward. And the other services, Space Force
and the Navy and Army and the Marines have adopted this same
type of thing.
And the other suggestion is the programs become--it was
referenced in a previous discussion, that, you know, the F-35
program slowed down significantly because of a problem with the
helmet. But it is because it was one giant program, and whether
it is a platform, an airplane, a ship, a tank, the plan form,
the platform that it is in is a development cycle of X number
of years, 8, 10, that they are good for that period of time.
They are 8- to 10-year, 12-year capability. The sensors, the
hardware in them you probably need to change out every three or
four years, in a plug-and-play, in an open systems
architecture, because the technology and sensor capability and
com capability changes that rapidly.
In the software area it should be a consumable. It should
be like petroleum, oils, and lubricants (POL), because you
change software so often, it is almost like the way you use gas
and fuel in an airplane, that you have to change it
continuously to stay up to speed.
So if there is a way to take a major defense acquisition
program (MDAP) and break it down so you are not one giant
program, that one flaw in either the software or a helmet or
one component slows the entire program down. Thank you, sir.
Chairman Reed. Thank you very much, General. Thank you,
gentlemen, for your insights.
Senator Cotton, please?
Senator Cotton. Thank you, gentlemen, for your testimony
today and your appearance.
Dr. Schmidt and Mr. Smith, I want to ask, to what extent do
your companies or, Dr. Schmidt, in your case, maybe the company
that you used to lead, or other companies that you may
represent, rely on Chinese suppliers for electronic hardware,
things like printed circuit boards, raw materials, like rare
earth elements? Dr. Schmidt, do you want to take that first?
Dr. Schmidt. I am no longer with Google.
Senator Cotton. Yes, I understand, but to the extent that
you still have knowledge of their operations.
Dr. Schmidt. Yeah. So, in general, the reliance is on
Taiwan, and I think that as a matter of national strategic
priority, Taiwan becomes more important to the United States
for that reason. The reliance on Taiwan is quite serious. I am
not aware of Mainland China dependencies, but there may be.
Senator Cotton. Mr. Smith?
Mr. Smith. Yeah, I think what we see at Microsoft is pretty
representative of what we are seeing across the IT sector for
hardware production, which is a pretty seismic shift towards
what I would call the diversification of the supply chain,
which means, frankly, moving more suppliers out of China and to
other countries. We are really focused on what I would describe
as a multi-country, multi-continent strategy, and what you are
seeing today is a lot of hardware manufacturing start to move
to countries like Thailand and Vietnam and Singapore. You
certainly see Taiwan, as Eric mentioned, as critical, South
Korea, Mexico, and the United States itself.
I think it is right to think of it in the following way.
The supply chain in China was created over the course of about
20 years, and I think with the exception of semiconductor
chips, where the fabrication costs are so high, we are probably
working through a transition of what I would call five years or
so, where you are going to see us and everyone else have a much
more diversified supply base.
Senator Cotton. Dr. Schmidt, you mentioned reliance on
Taiwan in particular. Are you talking about reliance primarily
on Taiwan for semiconductors?
Dr. Schmidt. Yes.
Senator Cotton. And that is an especially dangerous
reliance because Beijing considers Taiwan to be part of the
People's Republic of China. Correct?
Dr. Schmidt. That is correct, and if I may add that there
was a time when the United States was the great leader in
semiconductors, and indeed this Congress, in the 1980s,
approved something called SEMATECH, to make sure--it was
headquartered in Austin and was pretty successful in the eyes
of many people. But over the last 20 or 30 years, the majority
of the production of powerful semiconductors is now offshore,
with the exception of some of Intel's fabs and a few foundry
fabs.
But it is fair to say that if you want a leading piece of
hardware, which is what we all need to do what we do, you are
probably going to use a vendor called TSMC, which is the one in
China. They are just faster, better, et cetera.
One of the key recommendations that is in the AI report
coming out on March 1, is that America needs to stay two
semiconductor generations ahead of China, and that we need to
do the steps necessary to do that, which are long and
complicated and painful. But it is really important. We were in
this business. We got out of the business. We should back into
it.
Senator Cotton. Yeah. I just want to point out that our
dependence on TSMC is great, and the vulnerability of them to
China is great as well. Mr. Smith, you talked about South
Korea. There are some other countries you might diversify into,
in Southeast Asia. Those countries are still, let's just say,
within striking range of Mainland China, but China does not
have a core claim to want to forcibly seize their territory.
And this is one reason why Taiwan is not just a strategic and a
moral question for the American people but also vital to make
sure that we do not allow the Chinese Communist Party to seize
control of the world's most important chip manufacturer.
So I strongly support the efforts that we have to build
more semiconductors here in the United States. That is why I
worked with Senator Cornyn, Senator Schumer, and Senator Warner
last year on the American Foundries Act, and we are trying to
get money for it this year, but also to diversify, out of
striking range, let's say, from China, and in particular, out
of Taiwan itself. We want to be good partners with TSMC, and we
will always defend Taiwan's sovereignty and autonomy, but this
is not a vulnerability that the American people can continue to
permit.
Dr. Schmidt. If I could just add, there are quite a few
research efforts in America leading to new designs and new
approaches to semiconductor that would create the possibility
of a leapfrog. Those need to be investigated. That is part of
American greatness, and we need to emphasize them.
Senator Cotton. I agree, and Taiwan Semiconductor
Manufacturing Company (TSMC) is planning to open a plant in
Arizona. That is great as well. We want to help that kind of
reshoring of manufacturing of semiconductors as well.
My time has expired. Thanks, gentlemen, for showing up.
Chairman Reed. Thank you, Senator Cotton. Senator Shaheen,
please.
Senator Shaheen. Thank you, Mr. Chairman, and thank you all
for being here and for your testimony this morning.
You have all mentioned workforce as one of the challenges
that we face. Mr. Smith, do you believe we are producing the
science, technology, engineering, and mathematics (STEM)
workforce that we need to be producing to be competitive right
now, and how would you suggest we improve on that?
Mr. Smith. I think as we look to the future we are going to
have to invest more and we are going to need to do more, and it
really touches every aspect of not just education but skilling
for the entire population. I think we need to invest early in
the K-12 system. I think we need to support more teacher
training to get more computer science teachers into the
nation's high schools, in particular.
I think that our community colleges are an enormous
research that we are underutilizing today. There is an enormous
shortage in the United States today for cybersecurity
professionals, and I think we can harness our community and
technical college, and certainly our four-year and graduate
programs remain of imperative importance as well.
I do think we are also at a point in time where we should
think about providing people with digital skills as a life-long
endeavor. It means more investment in digital skilling for the
members of the military, but really every company, every
organization. And I think there are those of us in the private
sector--Microsoft has LinkedIn--we can do more and we are
working to do that, but it is going to require a collective
effort.
Senator Shaheen. I certainly agree with that. As a former
governor I know we worked very hard to focus on STEM in New
Hampshire. And one of the areas that we had some of the biggest
gaps were encouraging women, young women, to get involved. And
I think it is a place where Silicon Valley has not done a very
good job of providing equal opportunities for women. So we have
got challenges both in the public and private sector.
So do any of you have any thoughts about how we encourage
more women to----
Mr. Smith. I would first say we need to become more diverse
on every indicia of diversity. There is no area where----
Senator Shaheen. Absolutely.
Mr. Smith.--we should feel like we are ready to pat
ourselves on the back. We need to recruit and advance the
careers of more women. We need to do a better job of recruiting
and advancing more opportunities for black Americans and for
our Latinx population. And we should do it, I think, with the
recognition that our industry does its best work when we have a
workforce that is as diverse as the customers we serve, which
means the country as a whole.
Senator Shaheen. I certainly agree with that, and I think
the comments I think you and Dr. Schmidt both made with respect
to immigrants and trying to keep in the United States those
immigrants who have graduated from our colleges and university
with degrees that we need here is really important as part of
our national policy.
I want to go on to another topic, because I agree with the
sentiment that I think we all share, that China is the biggest
long-term threat to the United States. But as we look at what
we need to do to harden our digital infrastructure, clearly the
biggest recent threats come from Russia, and yet we are not
talking about how we combat that kind of cyber hacking into our
systems that are going to affect our ability to achieve all the
other goals that we have.
So do any of you have a thought about how we should be
responding to Russia and their cyber hacking, and what kind of
innovation we need in order to protect against those kinds of
hacks?
Mr. Smith. I would offer a few quick thoughts. Number one,
we need to modernize the information technology (IT)
infrastructure where it is dated, and it is often most dated in
the public sector. We are seeing this right now with vaccine
distribution and public health agencies across the country.
Number two, we really need to instill the broader
application of what are clear cybersecurity best practices. A
lot of these recent attacks have taken advantage of lapses in
just good practices.
Number three, we are going to need to secure the software
supply chain. We were talking before about hardware, but the
software supply chain, and really the build systems for
software need to be strengthened.
And then, finally, I would say we need to continue to
strengthen the rules of the road and hold other nations
accountable when they violate them, and do it with our allies.
Senator Shaheen. Thank you. General Carlisle, you mentioned
the SBIR program, which has been really critical in developing
innovation that has been adopted by the military. Right now
that program is scheduled to expire in 2022. How important is
it, do you think, that we need to extend that and make it
permanent?
General Carlisle. Senator, thanks very much for the
question. I cannot tell you how important we think that is, and
I think the ability to utilize SBIR, it is underutilized now.
It is another way that I believe, in reference to the
chairman's question of how we can accelerate the process. You
get an SBIR contract Phase 1, you maybe make it to Phase 2, and
you have a promising technology, but then how do you get it
into program of record? How do you cross that, quote/unquote,
``valley of death"? And there are different ideas about it,
whether it is a fund that allows you to put them into programs.
As a person that was the consumer, the problem we would
have is I would find this great technology and I would want to
put it into my F-22s or my F-35s, but I could not do it for two
years. A small business cannot survive two years on a promise.
You know, they are mortgaging their house so that they can make
payroll, so they can continue to develop this technology.
I think SBIR is incredibly important, and I think we need
to find a way in the authorization and appropriation process
and within the Department to have funds available to continue
those programs through Phase 3 and get them onto contract, and
more practical, use them in programs of record with the large
primes.
Senator Shaheen. Well, thanks very much. My time is up, but
if you have thoughts about how we should reform that program to
make it more effective for small business I hope you will share
that with us.
Chairman Reed. Thank you, Senator Shaheen. Senator Rounds,
please.
Senator Rounds. Thank you, Mr. Chairman. Gentlemen, let me
just begin by thanking you all for being with us today. Your
expert testimony in these fields is critical, and this
communication is very, very helpful to us.
Let me begin, I would like to begin with a question for
General Carlisle. Last year, the National Defense Industrial
Association, or the NDIA, sent our committee a letter stating
their concerns about the potential interference between the
proposed Ligado system and GPS. Last month, the Federal
Communications Commission rejected the National
Telecommunications and Information Administration's petition to
stay the commission's April 2020 Ligado order and
authorization.
What are your thoughts on the potential impact of Ligado's
proposal on the Department of Defense, and has anything changed
since the NDIA's letter last year? I think this is a critical
issue that needs to be addressed, because we are going to have
this come up time and time again in the future.
General Carlisle. Sir, thank you very much for your
question, and let me start by saying the work that the
chairman, the ranking member, and this entire committee has
done in support of the position that I believe is the correct
position with respect to Ligado cannot be overstated how much
we appreciate it. There are the two studies that go back to a
DOD study in 2018, and an Air Force classified study in 2016. I
was part of the Air Force classified study in 2016. I think
that it still stands. I think that the potential for
interference is great. They went from a space-based to a
terrestrial base, and knowing what the power of the global
positioning system signals are and the importance of position
navigation and timing, not just to the Department of Defense
but to the whole government and to every American, I think the
potential for an interference is something that has to continue
to be looked at.
I think we have to follow the science, and I think we have
to continue to study and learn. And until we really understand,
then I do not think I believe that you cannot move forward with
the Ligado until you finish the science and you actually know
and you can demonstrate that there is interference, or if there
is not then you can demonstrate that. But the risk of
continuing not knowing the answer to that and not having all
the science, I think that is unacceptable, sir.
Senator Rounds. Thank you. And for Dr. Schmidt and Mr.
Smith, what can be done to make sure that the Department of
Defense can maintain access to spectrum to meet warfighter
requirements while balancing the needs of the private sector to
build commercial 5G systems? Are there improvements to DOD's
related infrastructure that would help? Part of my question
also goes to being able to share the information, and sometimes
which is classified in nature, but to share the risks involved
when we have that challenge between commercial operations and
DOD, and the significance of the release of spectrum that may
very well be needed within the DOD's long-term plans.
Dr. Schmidt. About 12 years ago, the White House issued a
report, and I know because I was one of the authors, that
talked about the concept of preemption. And the basic idea is
rather than owning the highway you can occupy the highway, but
if a higher priority police person comes along you have to get
off the highway, or some metaphor like that.
So the way these systems work is the radio says, is this
busy with somebody who is more important, and if so then they
do not transmit. So this technology is now well mature and is
being used in something called Citizens Broadband Radio Service
(CBRS). I am one of the people who believes that we could share
the military spectrum such that the military had pre-emption.
That is, the military could always get what it needs but still
make that spectrum available when it is not used.
One of the more humorous example is that some of the
interesting key mid-band spectrum is using naval radars, and
the vast majority of those naval radars are not in the middle
of our country, on land. So you can imagine that there is an
opportunity to sharing. Anything that you were to do with
military spectrum would have to have an absolute rule that the
military had the highest priority, and further, I would propose
that the military run that sharing system to ensure it.
Senator Rounds. Thank you. Mr. Smith?
Mr. Smith. I would say, just building on what Eric said, I
think there is a broad recognition today that we are going to
need to use more effectively the so-called mid-band, between
3.10 and 3.45 megahertz, both for the DOD and for the civilian
sector of the economy. We are going to have to find a way to
share it, and I think he just offered a good description of the
kinds of approaches that have proven effective elsewhere.
And then we, you, are really going to need to decide what
is the best way to do that. There are two alternative models.
One has the DOD own it and then have others lease and operate
it. The other is to auction it and let the DOD have priority
access to it. I think that is an important discussion to have.
Senator Rounds. Thank you. My time has expired. Thank you,
Mr. Chairman.
Chairman Reed. Thank you, Senator Rounds. Senator
Blumenthal, please.
Senator Blumenthal. Thanks, Mr. Chairman, and thank you for
having this hearing, which is such a great way to begin this
session of the Congress, with a topic that is so timely and
critical.
First of all, let me say, on the semiconductor issue, this
shortage is real, urgent, and present right now. A group of us,
bipartisan group, wrote to the White House recently about the
shortage of semiconductors in the automotive industry, which
threatens to inhibit actual production right now in our
manufacturing of automobiles in this country. The same is true
in other critical sectors of the economy. I recently visited a
much smaller company, Sema4, in Seymour, Connecticut, which
produces medical equipment. It is affected by the shortage of
semiconductors as well. Its plea to me was, ``Please do
something to help us.'' So I thank you for calling attention to
this problem, but it is not some abstract future issue. It is
here and now.
Let me say to all of you thank you for your contributions
on the developing threats that we are discussing today. And,
Mr. Smith, in particular, I appreciate that Microsoft has been
such a leader in helping us to recover and understand the
recent SolarWinds attack. In fact, we are meeting here about
two months after the discovery of the largest cyberattack in
our nation's history, a devastatingly brazen and damaging
attack on our cyber defense, in fact, revealing the lack of
cyber defenses.
And I think that your reference to the recent crisis in
Texas shows us the mushroom cloud that, in the nuclear area,
would be the symbol of a similarly devastating attack in the
nuclear area. It is very difficult to sort of understand in
real terms what a cyberattack could do to this country unless
you look at what happened in Texas--loss of water, loss of
electricity. Our nation is in no way prepared.
So I would like to take your reference to the offense/
defense. You and I have discussed it a little bit. What can we
do to deter that kind of attack? Right now, we have failed to
make clear to our adversaries that they will pay a price, as
General Paul Nakasone said when he testified in his
confirmation hearing. He said our adversaries do not fear us.
What can we do either to make them fear us or establish, as you
put it, rules of the road that would establish some kind of
framework that will prevent this kind of attack on us or on
other nations?
Mr. Smith. It is a critically important question and, of
course, the ultimate answers will come from the people who lead
the government, not from those of us in the private sector. But
I would offer two thoughts.
First, it takes real clarity about the lines that others
cannot cross without consequences, because without that kind of
clarity I do not think any deterrent doctrine can be effective.
I am not even sure there is a deterrent doctrine in such a
situation. And I think it is easy to sort of lose hope that we
will ever bring the entire world together around new rules of
the road, but I do not think we need to. I think we need to
start with ourselves and bring our allies with us, and make
clear what lines we do not believe are crossable, and I would
say the disruption of the civilian supply chain, in a
disproportionate and indiscriminate way, should be one of them.
And then I think, like anything, there needs to be a
graduated set of tools. I think it needs to start by public
accountability with the United States and other governments, as
the country did in 2017, twice, after WannaCry and NotPetya.
But then there need to be responses as well, and there should
be a range of responses for different circumstances, but it
needs to be a robust menu, and we are going to need an
Executive branch that has the confidence and the support of the
American public to carry them out.
Senator Blumenthal. As yet there has been no response, at
least, that is known to us in the Congress. Maybe I missed that
response, either covert or apparent in some public way. There
has been no proportionate response, no response whatsoever that
I have seen to the SolarWinds attack, and I think that making
our adversaries, Russia, in particular, pay a price for this
attack is absolutely necessary. That is one of the ways to
establish some rules of the road.
But I agree with you that strengthening the supply chain
defenses is also important. And we have seen a wide variety of
competence in that area. For example, just in the government,
the VA has been much more defense-oriented, much less
vulnerable than, for example, the courts or the Department of
Justice. So we have seen varieties that I think we need to
learn from.
So thank you very much for your testimony today.
Chairman Reed. Thank you, Senator Blumenthal. Senator
Ernst, please.
Senator Ernst. Thank you, Mr. Chair, and gentlemen, thank
you very much for being here today. And, of course, as the
ranking member on Emerging Threats and Capabilities this is a
very, very important hearing for us today.
And, Dr. Schmidt, I would like to start with you. A number
of years ago I introduced legislation which became the National
Security Commission on Artificial Intelligence, which you chair
today, so thank you very much for that. And you did mention you
have a report coming out very soon on artificial intelligence,
and so maybe some of the questions I have for you today might
give us a little bit of a sneak peek on some of those efforts.
But as you know, and all of us understand, is that we have
a lot of different efforts across Department of Defense in the
area of artificial intelligence. So we have the Joint
Artificial Intelligence Center (JAIC), we have DARPA's
initiatives when it comes to AI, and, of course, then we have
our service branches and special operations forces all trying
to develop their own needs for AI to meet their requirements.
So a lot of different efforts coming from all different
directions, and, of course, that creates a challenge with the
coordination of those AI efforts.
So how is DOD working to make the different R&D centers,
the military branches and special operations forces efforts
available for AI development and those applications?
Dr. Schmidt. Thank you for giving all of us the honor to
serve on this commission. It has been a remarkable experience,
and I think you will be pleased the final report in a week.
With respect to your question, we recommended that the JAIC
be kept at a three-star level. In the military, hierarchy
determines everything, and it is important that it be at the
right level so that it has influence across the other
operations. The JAIC is well run. It does not have enough
resources.
In general, the way to understand the military is that
there are very few actual AI resources and there are an awful
lot of people who are attempting to help who do not know much
about it. And so we go over and over again the need for human
promotion, technical training, getting the right specialists in
the right positions, working with partners who are at the state
of the art. AI is extremely hard and confusing for a normal
programmer to understand, or a normal human to understand. It
is a new thing. It is very challenging. It needs specialists.
Senator Ernst. Yes, and thank you for that, and I think you
are right, in that we have many people attempting to take the
hill, and that is why the collaboration is so important with
the JAIC.
The current state of the AI strategy deployment at DOD, and
how, again, you know, talking about our near-peer adversaries,
how does this compare to the approach and the goals that have
been laid out by China with their own AI efforts?
Dr. Schmidt. It is hard to know what China is doing
internally. There is a classified report, which I obviously am
not going to mention now, that I would encourage you to take a
look at.
Senator Ernst. Thank you.
Dr. Schmidt. But a simple summary is that China has
announced that they wish to be the global leader in all aspects
of AI by 2030, and they are relentlessly focusing on that. They
are doing it with their STEM training, their investments, their
companies, and so forth, and presumably because of what is
called civil-military fusion, all of that information just
naturally goes back and forth within their military, unlike our
structure.
In the United States, we believe we are one or two years
ahead of China, not five or ten, and because of the diffusion
of the technology you have to expect that anything that is
invented in open-source AI world will immediately be adopted by
China. So the threat is very, very real.
Senator Ernst. Yeah. Thank you. And I think we should all
take note that, Dr. Schmidt, you said one or two years ahead of
China, and we cannot afford to lose that edge. And it would be
a much more comfortable margin to be five to ten years ahead of
China. So thank you.
If you had to prioritize, just very briefly, one or two
areas that would have an outsized impact at DOD when it comes
to AI at scale, what would those one or two be?
Dr. Schmidt. So when you speak to the senior leadership
what they want is a battlefield command center that takes all
the centers and helps them identify what to do. That should not
be the highest priority, because, one, it is hard, and two,
they do not have access to all the sensory data anyway because
they are all so stovepiped. So it is a good idea but do not do
that first.
Senator Ernst. Good advice. Thank you.
Dr. Schmidt. But it is important to say what not to do.
Most of the military spends most of its time watching things.
They watch for launches. They watch for cars. They watch for
aberrant appearances. AI and machine vision is particularly
good at that. An example is that I was on a minesweeper, which
is a wooden boat, where the young man who was doing it was
watching a screen to tell him--and his accuracy, by the way, I
asked his commanding officer, two-thirds of the time he found
the mine. Well, does that mean one-third of the time he
doesn't? Computers can do this much, much better, and plus the
guy is bored beyond belief.
So my point is vision, monitoring, and analyzing are the
best strategic uses of this technology--quickest to inform,
quickest to implement, highest payoff.
Senator Ernst. Absolutely. Thank you, gentlemen, so much.
The applications for AI are endless, and I thank you, Mr.
Chair, for bringing this hearing forward. Thank you.
Chairman Reed. Thank you very much, Senator Ernest. Senator
Kaine, please.
Senator Kaine. Thank you, Mr. Chair, and thank you to the
witnesses. I want to ask you about two topics. One is
immigration and the second is alliances.
So on the immigration side, just as in your industry, so
many of the most prominent advances in national security have
been innovated by immigrants or the children of immigrants.
Robert Oppenheimer, the Manhattan Project, child of German
immigrants. Jerry Jordanoff, who helped design the B-29,
Bulgarian immigrant. Father of the nuclear Navy, Hyman
Rickover, Polish immigrant. Father of stealth, Ben Rich,
Filipino immigrant. And then broader national security
priorities like vaccinations, Jonas Salk, child of Russian
immigrants.
How important is it if the United States wants to maintain
an edge in these emerging technologies, how important is it for
us in Congress to do comprehensive immigration reform that
continues to make the U.S. a destination of choice for talented
people from around the world?
Mr. Smith. I think it remains a very high priority. One of
the interesting things about technology is it always starts
with talent, so it starts with people. And if you want to have
the world's best technology, especially if you have a country
as we do, that has the world's best universities, you want to
continue to attract the best and brightest, not just to study
here but to stay here. And I think the more we can do in
especially these high-demand fields and these critical graduate
degrees, to give people the assurance up front that they can
not only get a visa but a green card, we put ourselves on a
path to do that.
I think one of the other reasons that comprehensive
immigration reform is so important is we have so many other
extraordinarily talented people here, including working in the
tech sector, who need the added certainty. They are either
stuck in a green card backlog because they came here from
India, and they risk actually having their children age out, or
they are dreamers. I am very struck. We have an extraordinarily
talented young person at Microsoft. He is working at Microsoft
to our benefit rather than on, frankly, what he would like to
do, which is the aerospace field, because as a Deferred Action
for Childhood Arrivals (DACA) registrant he can do one thing
but not the other. And so I just think we need to address this
range of issues to continue to nurture the world's best talent.
Senator Kaine. Let me ask Dr. Schmidt, if I could, about
alliances, and I would like to hear from others on this as
well, but to begin with, Dr. Schmidt. In your opening comments
you talked about seven areas where China is trying to get
dominance over the United States, where we are in competition,
seven technical fields.
My assessment, as a member of this committee and the
Foreign Relations Committee, is one area where the United
States still has some significant advantage over China is in
the area of alliances. We have longstanding alliances,
participation in multilateral organizations, and we do
multilateral alliances different than China does. China has a
little bit more of a mercantile, what-can-I-get-out-of-you
approach, and the countries seem to understand that. And it
does seem like adversaries like China and Russia, to the extent
that they are nervous about us, one of the things that most
makes them nervous is alliances like NATO and others, or when
the United States was leading, potentially, into the Trans-
Pacific Partnership (TPP). That made China very, very nervous.
In the area of emerging technologies, how can we use our
alliances to help us drive an expanded capacity without running
into a problem, say, for example, the F-35. Built it with
allies, Turkey has been sort of a wavering ally, and then we
end up building something, and there is a security compromise
as the technology now is available to a wavering ally. How can
we leverage the value of alliances in advancing in these
emerging technologies while protecting ourselves from an
example like I just made with the F-35?
Dr. Schmidt. Thank you, Senator. I note with concern that
Boris Johnson announced today that they are all Sinophiles and
that he is heavily motivated to work with China. This is our
longest-standing partner, the United Kingdom. This is a bad
sign and a bad omen for what is going to happen. We must build
every possible technological sharing path between our key
alliances, and who are they? Israel, France, Germany, the
United Kingdom, Japan, Korea, maybe India. There is a list of
about ten. The word that is coming to the industry is the T-10.
And what it means is constant harvesting of the best ideas,
putting companies together, and so forth.
If you start from my premise that American global companies
are our greatest asset because they move so quickly, let's have
American companies working closely across all those boundaries.
Everything that we do to make it harder to work across those
boundaries also hurts our national security.
I also think that the government should have a national
competitiveness plan, which includes a list of the key
technologies and a list of the key countries. There should be
money--not a lot of money, but basically money to basically
fund the communications, travel, and the partnerships, with
somebody driving it out of the White House.
Senator Kaine. Illuminating answers. Thank you. Thank you,
Mr. Chair.
Chairman Reed. Thank you, Senator Kaine. Senator Cramer,
please.
Senator Cramer. Thank you, Mr. Chairman, for having this
really impressive panel and hearing. You all have drilled down
pretty deeply on several issues that I have an interest in, so
I am going to try to drill just a little deeper on one, first
of all. It gets to what you said, Mr. Smith, in talking about
procurement reforms. I do not know that we could disrupt enough
to be as effective as we need to be, but we, in this
exceptional system of ours, protect things a little more
probably than other places, and that is okay.
But you specifically raised reform of the protest--you
talked about some protest reforms. Could you elaborate a little
bit on that, because I agree. That is a problem. You have all
talked about the delays that lead to delays, and time leads to
mischief--those are my words, not necessarily yours--but
protest reform seems to be one of those areas maybe we can do a
little better while still protecting everybody.
Mr. Smith. Well, it is a really important question. It is
certainly another one that we have experienced as a company
over the last year.
I would start with the recognition that these new
technologies that we are talking about today really, for the
most part, start as commercial technologies and then they are
put to military use, rather than the other way around. So the
best way for the Defense Department to move faster is use
commercial technology, add security layers, as we have done
with the DOD and the intelligence community for, say, secret
and top secret workloads, and then create adaptations. But it
is so important to move quickly. Then the question is, how do
you move quickly when the protest process moves slowly?
So I do think there is a real opportunity to look at the
process, streamline it, put in place some tighter deadlines,
consider legal reforms that would apply those deadlines to the
judicial aspects as well. We do not think that others should be
denied an opportunity to protest. Maybe for better and worse
that is part of the American way, to some degree. But it sure
would be beneficial if it could move faster.
Senator Cramer. Others on the same topic, Dr. Schmidt or
General?
General Carlisle. So the only thing I will tell you I
noticed, and Mr. Smith and I had this discussion ahead of time,
is I agree, there has to be an opportunity, but the speed with
which you go through it--and the fact is there is absolutely no
disincentive to protest. And except for the consumer, the
customer that is going to actually use the equipment and is
denied that equipment for an extended period of time. So the
question is, how do you accelerate that, allow those things to
happen fairly, but at the same time do not disadvantage the
person that is waiting for the equipment while you are waiting
for the protest to be resolved.
Senator Cramer. For sure. Well, I would love to drill down
more on that, if anybody has any brilliant ideas, whether it is
our judicial system, legal system, regulatory system, or
government, whatever we can do.
You also have all talked a lot about the skilled workforce,
and I think you have answered a lot of the questions really
well on that. One area I might just seek a little more input
on. First of all, I agree wholeheartedly. We have so blown the
opportunity to maximize the incredible high-skilled immigrants
that have come to this country, whether for education or for
work, or all of the above, putting them in these boxes. The
backlog of green cards is immoral to me. The per-country caps I
have been trying to get rid of for a long time. It punishes
certain countries, obviously, that have a lot more to offer us.
But it also opens up another one of those security risks,
right, I mean, whether it is chip manufacturing or immigrants.
How do you see moving forward with high-skilled immigrants and
some of the reforms, whether it is--I think you have talked a
little bit about comprehensive, and comprehensive is fine, but
comprehensive seems to always get in the way of doing some
other good things. And I am just looking for lane here in this
next Congress to finally get something over the top as it
relates to the backlog of green cards and high-skilled
immigrants.
Dr. Schmidt. So Brad and I have spent 30 years here saying
basically the same thing.
Senator Cramer. Well, good. I feel better. I have only been
spending about six.
Dr. Schmidt. I know, and I am sorry to say the same thing
again. Our industry is critically dependent upon high-skilled
workers. Today, our industry represents 20 or 25 percent of the
total stock market value of America. So we are sort of
important in at least the economic output, if not the pride of
the country. And we need these people because they are the
creators of our products.
What I would suggest with respect to the questions of
concern about security is that you could imagine, for example,
a Chinese national comes in, and you would ask them, ``Have you
ever been associated with this group, this group, and this
group?'' and presumably they would say no. When you discover
that that is the alternative truth, through some mechanism, you
can get them out. And I think that there is an investigative
process that is relatively straightforward. There is set of red
flags. The vast majority of the Chinese people that we work
with are not political, not dangerous, and they are incredibly
important.
One more comment. We looked at the question of how
important are Chinese researchers for the AI effort in our
report, and it turns out that the Chinese researchers are the
number one foreigners on the key papers. So if you were to, if
you incorrectly get rid of all of them, because you just do not
like them or something, you will, in fact, hurt America's AI
leadership.
Senator Cramer. Well, I might--as I just wrap up with my
time gone--submit to you as well that you have discussed allies
and alliances, and this is another area of opportunity, it
seems to me, to build maybe some new alliances with some large
countries. And with that I yield. Thank you.
Chairman Reed. Thank you, Senator Cramer. Next will be
Senator Gillibrand via WebEx. Senator Gillibrand?
Senator Gillibrand. Thank you, Mr. Chairman. Thank you for
testifying. Since Cyber Command unified the cyber defense of
our nation in 2010, we have adopted a strategy of persistent
engagement, which intends to keep our adversaries continually
challenged in order to stop attacks like this before they
begin. The SolarWinds attack has been going on for nearly ten
months and was likely designed by over 1,000 software
engineers. What resources do you believe that we need to
develop in order to avoid missing something like this again?
Dr. Schmidt. Can I add, Senator--can I add that the
vulnerabilities of the nation's infrastructure are well known
and we have chosen not to fix them. If we wanted to fix them we
would upgrade all the software and we would have some rules.
So, for example, the data that is inside these systems is
encrypted at rest. It is encrypted in transit. We would use
proper authentication keys. The military actually does this.
Many of the rest of the aspects of the Federal Government do
not.
So until we commit to bringing our infrastructure up to the
state of the art of defensive tools we will continue to have
this exposure, independent of what United States Cyber Command
(CYBERCOM) does.
Senator Gillibrand. Right.
Mr. Smith. And, if you want, I would add just two quick
responses to your question. One is the recent attack exploited
the fact that while the National Security Agency (NSA) has
authority to look outside the United States, it does not inside
the United States, and it was, in fact, it appears, data
centers of commercial companies in the United States that were
used really for much of this activity. So I think the Congress
and the country are going to need to decide how it wants to
better protect our internal resources.
And then second, related to that, I think there is a real
question, when must companies, under the law, a law to be
decided, report these kinds of attacks, and to whom and how in
the government? I think we need to consider how these things
fit together so we have more aggregated and comprehensive
threat intelligence.
Senator Gillibrand. Thank you. On January 6th, we saw what
can happen when extremism, incubated in online social groups,
spilled over into the real world. Many hate groups, present at
the Capitol insurrection, used online platform to organize and
rally. The development of emerging technologies, including
improved encryption and other communications tools, are a boon
to the privacy of our citizens but also obscure already murky
online extremist networks.
What responsibilities do you believe private industry has
to disrupt the spread of violent extremist ideology, and what
are the possible regulatory changes that Congress should make?
Mr. Smith. I think this really goes to the question of
addressing harmful and dangerous content online. If you look at
the trend around the world, you know, we have seen other
governments take this on. Australia was a leader a couple of
years ago in enacting new legislation, that imposes
obligations, legally, on tech companies, including, you know,
Microsoft, Google, and others, to address extremist violent
content and terrorist content.
As an industry, we have moved to work more globally and
beyond the law, in a collaborative way, through what is called
the Christchurch Call, which has brought together a number of
governments and the leading tech companies. We are doing more
to address this. I do think this is a moment in time when we
should ask where we want the law in the United States to go and
where we want collaboration with our allies to go.
The U.S. work is always more complicated, frankly, than in
other countries because of the nature of the First Amendment to
our Constitution, but a lot of these efforts have identified
weak points we can work together to address.
Senator Gillibrand. Thank you. Just one last question on
China before my time expires. Obviously, China is becoming, and
aiming to become the global science and technology leader by
2049. How can we best prepare to outpace China? What obstacles
do you see the U.S. having to overcome in the science and
technology race? I did hear your testimony about software and
the importance of investment and collaboration. What do you
believe are the biggest missteps to date, and what do you think
are the best ways to avoid it in the future?
Dr. Schmidt. My personal view is that our industries'
success has largely been due to the extraordinary decisions
made by this body over 50 years to fund basic research,
starting with Vannevar Bush, et cetera, et cetera. Today, R&D
funding, as a percentage of gross domestic product (GDP), is
lower than it was at Sputnik. So one of the problems is that we
are, to some degree, leading off of our seed corn, if you will,
on all of that. We have already talked about immigration. We
have talked about the importance of STEM education, and those
things.
I think we have to confront the following problem. There is
a set of platforms, which I identified in my technology, which
are going to happen but they are going to happen first in
China, unless we have a more concerted effort in America. I
would like to see a national list of key technology platforms
that we collective agree must emerge, must emerge using Western
values, must be the ones being used by our partners.
And to understand what happens if we do not do that,
consider Huawei, which we are basically trying to ban as hard
as we can, because their products were less expensive, more
easily subsidized, and faster, in some cases, than the
competitors that are from Europe. America got out of that
business. That is an error. I want us to be in those businesses
with world-class products. I think we need to know what that
list is, I think the government will need to help with some
forms of funding, and we need to let the private sector build
those things and make it successful.
Senator Gillibrand. Thank you, Mr. Chairman.
Chairman Reed. Thank you, Senator Gillibrand. Senator
Sullivan, please.
Senator Sullivan. Thank you, Mr. Chairman, and, gentlemen,
thank you for testifying today and your service. General
Carlisle, always good to see you, sir. Great career in the
military. And, Mr. Smith, I appreciate our opportunity to chat
last night. It was very informative for me. And, Dr. Schmidt,
thank you for all you are doing in your post-Google world.
Let me ask a question. There has been some press back and
forth, and I think given you three leaders, right, big leaders
in America, in a whole host of different ways, there has been
some press in the last couple of years where some concerns I
have read, and I would love you to just comment on it, our tech
industry, Silicon Valley in particular, kind of maybe not being
so interested in supporting our military, supporting the
Pentagon.
You know, Mr. Smith, as you and I talked about it, I had
the opportunity to go out to Microsoft's Integrated Visual
Augmentation System (IVAS) partnership and development center.
I thought it was incredible seeing these young men and women
who were very motivated to help our men and women in uniform.
Dr. Schmidt, I am sure you saw, there is some Google press that
I thought was very unfavorable, where, you know, there was this
idea, hey, we do not want to help the Pentagon. My view, as an
American, it is a free country. You can do whatever the heck
you want, but do not then be found to be helping the Chinese
Communist Party. Like that is going to be a problem.
So can you three--I would love hear just succinct
statements on, from your perspective, just how important that
is. We have a challenge with this very new, great power
competitor and the technology aspects of our country. Working
with our military is going to be indispensable. And it does
concern me some when you hear--and again, they are allowed to
do it; that is one of the great things about our country, it is
free. You can say whatever you want. But I would love to hear
from you guys on just how important it is to be doing what you
are doing and what we are talking about here, because if we do
not have that kind of cooperation it is going to be tough on
all of us.
Mr. Smith. Well, I would say first I think one of the great
challenges for this committee, the Congress, and the country is
to keep the public united around the importance of our national
security at a time when we live in a polarized political
climate. And the key to that, not surprisingly, is, as always,
leadership and communications.
The formula that we have found to be effective is to be
clear, that we, as a company, at Microsoft, will provide to the
United States military all the technology that we create. We
will simultaneously engage to address the issues that a new
generation I think rightly focuses on, things like the ethics
of artificial intelligence. And we will honor people's choices,
and when we have a project like IVAS it is really an all-
volunteer project, and we have no shortage of volunteers.
Senator Sullivan. And those young men and women, I will
tell you, having spent a day with them, were incredibly
impressive, motivated, patriotic, because they knew what they
were doing, which is helping the frontline troops who put their
lives on the line for our nation.
Mr. Smith. And I think there is one other thing where
communication can be invaluable. Look, most people in the tech
sector or perhaps most industries are simply not aware of the
deep ethical tradition of the United States military. And when
they learn about it we actually realize that we have more to
learn from the military, and it really changes the climate
among especially a new generation of employees.
Senator Sullivan. I appreciate you saying that, Mr. Smith.
Dr. Schmidt or General?
Dr. Schmidt. So the only thing--I am sorry, sir. I did not
mean to interrupt. The only thing I would add is, you know, my
experience is the American population is further and further,
in many cases, removed from the military. It is an all-
volunteer force, which is exactly the right thing, in my
opinion, and the quality of the force in the United States
military, I tell everybody if you want to be impressed with
America's youth, go out to your flight lines, your ships, your
tanks. These 19-year-olds are just amazing.
Senator Sullivan. It gives you hope and optimism. There is
no doubt about it.
Dr. Schmidt. But I think it is an education. I think that,
just as Mr. Smith said, I think, you know, a lot of it is not
because they fundamentally, you know, do not like the military.
They just do not know the lengths we go to to deter and
prevent--the last person that wants to go to war is the person
getting shot at. And so the prevention and deterrence. And then
when we are to follow the most ethical rules, if we have to
engage and how we engage and how we do everything we can to
only follow the enemy combatant.
So I think it is an education process, Senator Sullivan.
Senator Sullivan. Dr. Schmidt, do you have a view on that?
And I did not want to come down too hard, but I have ripped
some Google employees in hearings like this before, where maybe
it was bad press reporting, but I was like, you have got to be
kidding me. I mean, again, free country. You can do whatever
the hell you want. But if you are not going to help the
Pentagon please do not go help the communist party of China
with their AI research.
Dr. Schmidt. I did not agree with the Google decisions on
Maven, et cetera. As you know, I worked as a government
employee, working for the DOD for five years, using the DIB, so
my personal view is clear. I also funded and have continued to
work with a large number of startups in the areas that we are
interested in, who are really, really committed to working with
the DOD. So I can tell you that the Google experience you had
was probably an aberration compared to the industry as a whole.
Senator Sullivan. And, Mr. Chairman, if I may, just very
quickly, since this is such a distinguished panel--sorry to my
colleagues--but I know you have been getting a lot of questions
on China. Just very quickly, to be respectful here--I am over
my time--comparative advantages that we have versus what they
have, particular in this tech sector. I mean, I will give you,
I think the fact that we are an energy superpower right now,
producing more oil, gas, and renewables than any country on the
planet, China would love to be in that driver's seat.
Unfortunately we have an administration right now that wants to
diminish that, which I find ridiculous and crazy.
But where do you think the comparative advantages are,
particularly in AI? I read that part of their advantage is
their massive population, that in some ways their own
population is guinea pigs that helps them advance in AI. Where
are our comparative advantages, and vice versa, theirs?
Dr. Schmidt. So the Chinese are well ahead in areas like
face recognition, because of what they do to surveil their
citizens.
Senator Sullivan. So that is the idea of guinea pigs and
billions of people that they can just test it on?
Dr. Schmidt. Their technology is generations ahead of what
is possible in the West, and you can understand why. Their
technology is extremely far ahead in electronic commerce and in
mobile payments, and most recently they have announced the
development of a central bank digital yuan, their currency, to
actually--and they obviously have, from their perspective,
internal security benefits from watching where all the money
goes. These are all things that the United States would not do.
So those are two where there is no question that they have an
advantage.
There are people who believe that because they have
essentially no privacy rights, in the terms that we think of
it, that they will be able to aggregate very large databases,
in particularly in health care, and that will allow for them to
discover new things and so forth. We need to address these, and
again, without compromising our core American values.
Mr. Smith. I would just say, very briefly, we often talk
about research and development, but especially for something
like AI we should talk about research, development, and
deployment. In other words, broad adoption and use, especially
when you think about the positive feedback cycle that is
created when technology is deployed. It creates more data. That
data then leads to further improvement.
I think China is doing a better job right now than we are
in deployment. Part of it is it is government-led in many ways.
Part of it is there are uses where we, quite rightly, say no.
Part of it is the entrepreneurialism we are seeing in many
parts of the Chinese economy. So I think for the United States
we have to think about how we foster faster deployment, and I
think in the government, for the DOD, how the DOD, for example,
can foster faster deployment.
Now, at the same time, the American comparative advantages
in other respects remain considerable--our universities, our
commercial technology sector. And I think the principles. One
thing we have not noted that I think is very important in the
world today is the fact that the DOD, last year, adopted
ethical principles to guide the use of artificial intelligence
by the military. And I think the more we can encourage our
allies to adopt these principles, the more we separate
ourselves in a way that will benefit us in numerous respects.
Senator Sullivan. Thank you, Mr. Chairman.
Chairman Reed. Thank you, Senator Sullivan. And now via
WebEx, Senator King.
Senator King. Thank you very much, Mr. Chairman, and
welcome to our distinguished panel. We have touched on a lot of
important issues. Let me start with a little bit of a detailed
question for Mr. Smith from Microsoft.
You touched upon this. It strikes me that we have a gap in
our authorities towards detecting and dealing with cyberattacks
in that the Central Intelligence Agency (CIA) and the NSA are
restricted from operating within the borders of the United
States, and yet the attacks, like SolarWinds and more and more,
our enemies are getting more sophisticated about using servers
within the United States. It leaves the FBI as sort of the de
facto only cyber defense. Am I correct that is something that
we really need to look at? We do not want to be spying on our
citizens. On the other hand, we do not want to leave ourselves
defenseless. Brad, your thoughts on that?
Mr. Smith. Yeah, no, Senator, I think it is a really
important question, and I think the first question for the
Congress and the Executive branch is what part of the
government do we want to have assume responsibility for what I
will call the aggregation of threat intelligence domestically.
Is it the Cybersecurity and Infrastructure Security Agency
(CISA)? Is it the Federal Bureau of Investigations (FBI)? Is it
somebody else?
The FBI, obviously, is principally responsible for law
enforcement, which means it can work with the Department of
Justice (DOJ), it can use its subpoena power, but, you know, it
then needs to protect the confidentiality of information to
investigate a crime. And what we are really talking about here
is threat intelligence information that needs to be shared
rapidly, oftentimes immediately, with the other parts of
government.
So I think this is a key question. What part of the
government should do it? What should the process be for
collecting it and for sharing it?
Senator King. Great. Thank you. Dr. Schmidt, an additional
question on a different area, and you have really touched upon
it today. Industrial policy has a bad name in this country but
that is really what China is engaged in. And you mentioned we
used to do a lot more R&D, we need to establish priorities, we
need to bring semiconductor manufacturing home. Are we really
talking about some kind of at least a more pragmatic and
planned attack on maintain the technological edge? Is it
Industrial Policy 2.0?
Dr. Schmidt. Senator, I hate to say yes, it is industrial
policy, but can we not call it that? I think what would be
useful would be to say there is a set of things that have to
happen in America to maintain leadership globally in the
important areas, and remember, these are the technologies that
drive all of our economic output, our global presence, and so
forth, and we need to do whatever it takes.
I think in many cases, with a little bit of focus, with a
list, with leadership from the White House, leadership from
here, a set of gatherings, and so forth, we can agree on what
to do, and it is not as much the money as it is getting all the
forces aligned.
What I learned in working on your AI report is there are
plenty of people doing a lot of things, and they are somewhat
discontinuous. And getting them unified around five or six or
seven activities would be very helpful. In particular, we have
highlighted--Senator Cotton and others have highlighted this
question about semiconductors. That is a key issue. How are we
going to solve that problem? Let's get some people in a room.
Let's try to figure out what is the fastest path. If they cost
$50 billion and it works then maybe that is the right tradeoff,
but I would like to have that debate.
Senator King. Thank you. One final question, again for Brad
Smith. I went to a defense policy conference in Singapore three
or four years ago, two or three years ago, and met with a dozen
or so officials of a variety of Asian nations. I came away from
that with the conclusion that we have allies and China has
customers, and that most of those countries wanted to work with
us but they were always looking over their shoulder at China.
In terms of cyber defense, in terms of national defense, in
terms of technological innovation, it seems to me that allies
are one of the most important assets that we have, that really
most other countries, and particularly our adversaries, do not
have.
Mr. Smith. I think that is very well put. One of my
favorite publications every year is the January edition of The
Economist. It is an assessment of the world's democracies by
The Economist intelligence unit. This year it says that there
are 75 democracies in the world. They account for 49.4 percent
of the world's population, roughly half of the world's people.
And what it also notes this year is that democracy is growing
in a number of important countries in Asia.
And I think it is a powerful remainder for all of us that
there is an alliance of the world's democracies that we need to
nurture as a nation, that we need to invest in and support as a
technology sector. And we do that well, it not only advances
the values that we all support in this country, it makes our
technology base stronger. When you pull together these
countries, you do not even have to pull them all together. Eric
was talking about this before. But when you get India together
with NATO and countries like Japan and South Korea and the
like, and you pretty quickly get more than 2 billion people,
that is a bigger market, obviously, than China.
Senator King. And it is also a huge aggregation of talent--
--
Mr. Smith. Absolutely.
Senator King.--that can be taken advantage of.
I will leave you with a thought from Churchill. You can
never miss with Churchill. He said, ``The only thing worse than
fighting with your allies is fighting without your allies.''
Thank you very much, gentlemen.
Chairman Reed. Thank you, Senator King. Senator Tillis,
please.
Senator Tillis. Thank you, Mr. Chairman. Thank you,
gentlemen, for being here. I am sorry that I was not here. I
have been watching it on TV and participating in two other
committees that are meeting simultaneously. But I was here for
your opening comments.
One thing that, as I was reading the committee prep
materials I was thinking we need to do differently is how can
we really accelerate the pace of innovation within the DOD for
our defense. And I went back to Operation Warp Speed. Are you
all familiar with that? We made, in record time, innovated a
vaccine, did a public-private sort of bet on people in the
private sector who were willing to take the risk, but the on
the back end had Federal funding available for them if they
produced a result in a shorter period of time.
Do you think if we are really going to accelerate, break
through some of the--Mr. Smith, you and I talked last night
about some of the hurdles that we have in DOD to just
accelerate and field technology--should we be thinking about
innovative ways of preparing or moving up to the NDAA to really
incent more private risk-taking with some federal backstop,
based on specific outcomes? I can think of a number of specific
areas, but does that make sense? Is that something that a
Microsoft would look at?
I want to go down the line. We will start with you, Mr.
Schmidt, Dr. Schmidt.
Mr. Smith. Yeah, I think it is an excellent question and
there are two thoughts worth considering. Look, first, any time
we can have more risk-taking in the private sector that is a
good thing, and not every company can afford to do it.
Microsoft can do things that a small businesses cannot. But
look, we built a manufacturing facility in Milpitas,
California, for our IVAS goggles for the Army before we won the
contract with the units that we would produce there. That was
private risk-taking.
We have literally been frozen by a Federal court on our
performance under the Joint Enterprise Defense Initiative
(JEDI) contract for more than 12 months. We have never stopped
working on it, not even for one day. We may never get paid.
That is a risk we are running. The customer may never be able
to use what we create, but we have the confidence that what we
are building will be of benefit to the United States some way,
somehow. So the more we can encourage private risk-taking I
think is a good thing.
And then, specifically, I do think there is something to
think about in terms of lessons from Warp Speed for certain
areas of technology. If you think about quantum computing,
there are some that think it will take 20 years. There are some
people that think it will take a decade. A year ago we were
debating whether it would take 10 years to get a vaccine, and
it took less than 12 months. And it did benefit from government
spending, putting some money behind a series of companies with
different techniques. Do not bet it all on one company or one
method. Prepay and do it on the basis of particular milestones,
so the government is getting in advance what it would then own
or be able to use if something crosses the finish line.
But, you know, there is something there, I think, that we
have all learned that sort of surprised us, I think, in the
last year, that we should now apply to some of these new
fields.
Senator Tillis. Dr. Schmidt?
Dr. Schmidt. I agree with Brad. I would recommend that in
this year's NDAA you all identify four projects where you say
they will be run radically differently. I would pick one in
missiles, one in satellites, one in personnel, and another one
in some other areas. And you would, by law, state that they
will not be run using the normal procurement mechanisms, but
rather you will appoint a joint committee from the Congress as
well as the Pentagon and give them the freedom to run the
experiment.
Senator Tillis. And General Carlisle, I am also thinking
about the reality is some of the most brilliant ideas may come
from some of the smaller players that are virtually impossible
for them to do, just because of their scale with the DOD. But
do you think that that concept would apply with the right
portfolio of some of the smaller companies? That is what I have
in mind. The big players have to be there because they have the
scale, but how would we structure that, I think building on Dr.
Schmidt's suggestion for the NDAA. I honestly believe we have
to have accelerators like this if we do not want to be talking
about this next year when you come back.
General Carlisle. Yes, so I could not agree more, Senator
Tillis. You know, I think the Department has got to be willing
to take risk. It is risk averse. If you are a program manager
in acquisition or a contracting officer you do not get promoted
because you took risk. You get promoted because you are on
cost, on performance, and on schedule. So you do not try to get
a stretch goal on performance, and that is where innovation
comes from. You do not try to get it faster, because you may
not make it. So we have to figure out how to incentivize inside
the Department and industry. And I think your point on, you
know, what we talked about earlier with Senator Shaheen, is the
Small Business Innovation Research fund, we have got to find a
way to get those through the tough times of an extended
process, make it faster, and then allow them to be able to stay
competitive and bring those technologies to the warfighter.
Senator Tillis. Thank you, Mr. Chair.
Chairman Reed. Thank you, Senator Tillis. And now via
WebEx, Senator Duckworth, please.
Senator Duckworth. Thank you, Mr. Chairman. Gentlemen, I
apologize. I am having a little trouble with my video, because
of bandwidth, but I am going to go ahead and do this via audio.
Thank you so much for your testimony today.
The entire DOD has to innovate to compete against the other
great powers, but U.S. Transportation Command (TRANSCOM) faces
a unique set of challenges. Transportation Command's
communications network, systems, and software have to support
deploying troops and sustainment around the world. They receive
inputs and data from many different government entities and
also via doing business with private companies, for example,
shipping companies and commercial air carriers.
But cybersecurity vulnerabilities in Transportation
Command's network risk risks exposing our troops' locations,
readiness levels, and operational plans, and the requirement to
work with private business complicates addressing these
weaknesses.
Dr. Schmidt, during your time on the Defense Innovation
Board, the board produced a number of recommendations regarding
the DOD's digital networks and cybersecurity vulnerabilities.
In your opinion, how should Transportation Command, in
particular, approach rapidly improving its cybersecurity
without losing its ability to respond to warfighters and work
with civilian entities? Your suggestions could include
technical innovations, organizational changes, or perhaps
policy proposals, for example. And I love this idea of picking
several projects and approaching them radically differently in
terms of procurement. Thank you.
Dr. Schmidt. Thank you. So our group actually visited St.
Louis and the Transportation Command it was a very, very
interesting visit. The key room is the room where you have
people in uniform who basically have two screens, and there is
an order from one shipping system and they type the number of
the order into the other screen and cause it to move along. So
that is the level of automation that we, unfortunately, have in
that. Any company would have integrated that, and we
recommended that.
My own view is that there is a proposal in Transportation
Command to do a new transportation system, which was hung up in
a bunch of procurement issues. But the 80 or so different
systems are going to have to get replaced by a more unified
system, and that more unified system will have to have modern
security. That is how we would address your concern. Because of
the way it is currently architected, you are correct that we
are very exposed to attack because there are so many different
systems that are disparate and they are not unified.
Senator Duckworth. Thank you. General Carlisle, do you have
any recommendations, based on your work with the commercial
members of the National Defense Industrial Association?
General Carlisle. Yes, ma'am, and, Senator, thank you for
the question. I agree with Dr. Schmidt, and I think we saw it
in the command centers as well and how we integrate across
different systems, even jointly between the services. And I
think, you know, the comment was made earlier. We have a
tendency to have our sensor suites are all stovepiped and our
communications are often stovepiped. And what industry needs is
the common architecture and the ability to work across the
different systems, and I think Transportation Command is a
great example of that, where they are working with the whole of
government, really, and the commercial enterprise, but the
systems are not compatible.
So what Dr. Schmidt said, and our ability to drive industry
to have a set of standards and out of the stovepipe challenges
that we face today in many of our systems as they try to
communicate.
Senator Duckworth. Thank you. Gentlemen, I am closely
watching the progress of future vertical lift, mostly because I
am personally interested in advancement of rotary-wing
aviation, as a rotorhead myself, and also because the Army has
made a number of smart decisions as it has developed a program
now. I am hoping some of these decisions can be adopted across
the DOD [inaudible].
Chairman Reed. You broke up, Senator Duckworth. If you
could repeat the question.
Senator Duckworth. Okay. I am going to turn my video off,
because that seems to be the problem here. I apologize.
I was talking about the future vertical lift, and language
I had in last year's NDAA requiring a review of lessons learned
and employing open systems architecture in the FVL program. Dr.
Schmidt, what are the benefits of using open systems
architecture in programs like future vertical lift (FVL), and
what barriers do you see to the military services using this
approach in future acquisition?
Dr. Schmidt. Thank you. I love your question because I am
also a very big helicopter person.
Senator Duckworth. Fantastic.
Dr. Schmidt. If you look at the way the aviation world has
worked, many of the structures and so forth are relatively
secret and proprietary. And what we have learned with more
sharing across the industry, the whole industry moves faster.
So I strongly recommend that open source designs be made
available. And my personal view is that the way the Defense
Department should do these things is that the Defense
Department should have design studios that design things which
are owned by the government, and then that technology that they
own is then given to the manufacturers to then develop further.
But I would like the government to own much more of its own
intellectual property by developing it itself, by funding
teams, design teams. I also think that that will allow for
faster iteration throughout the primes and their manufacturing
cycles.
Senator Duckworth. Thank you. And I am out of time, but if
you could follow up with any type of barriers and any
recommendations on overcoming barriers, in written form, after
the hearing I would appreciate it. Thank you.
Chairman Reed. Thank you Senator Duckworth. Senator Scott,
please.
Senator Scott. Thank you, Chairman. First off, I thank each
of you for being here.
General Carlisle, you recently retired. In the roles you
had in the military, how concerned were you about, you know,
what technology companies were doing, I mean, the theft by
Russia and China of technology, the espionage, things like
that, and did you feel like you were at a disadvantage as
compared to what Russia and China military was doing?
General Carlisle. Sir, we have the greatest fighting force
and the greatest military in the world, and I believe we have
the greatest equipment in the world. Some of the programs that
I was in, that are now declassified, I was part of the
exploitation of some of the capabilities of our adversaries,
both the Soviet Union (USSR), at the time, back in the late,
great days of the Cold War, and China. And, by far, our
equipment is superior to our adversaries. And you can tell that
not only from what we got to see but our friends, partners, and
allies want to use our equipment as well, because of the
quality of it.
I do believe that gaps is knowing because of the theft that
occurred. I was in China when I was the commander of PACAF, and
we were walking up and down the line looking at their
airplanes. I actually got to crawl into a couple of their
airplanes, a J-10 and a J-12, and when you looked inside you
could tell that it was just--they took stuff from wherever they
could steal it, to put it in those airplanes. And the result is
that the gap we had, the superiority we had against our
adversaries, because of intellectual property (IP) theft,
course of action that I talked about in my opening statement,
that gap is narrowing. And that is why we have to continue to
get innovation out more quickly, because in today's world you
just do not maintain----
Senator Scott. But then what you just heard, what Dr.
Schmidt just said, that we do not even have systems that--you
know, you had to put something from one system to put
information into another one. I mean, in real time you are not
going to win a war if you cannot do some basic things like
that, where we do not have the ability to share information
rapidly. You know, it just seems to me that we have not used
the private sector, and we do not have the relationship with
the private sector, for whatever reason. But China does, and
China might because they steal it, but they do have, you know,
whether it is AI or things like that, they are going down a
path that we are not even--we are going awfully slow in.
General Carlisle. Senator, you know, I do not disagree with
that. I think that is a challenge as we move forward. We do
make it work, though. I mean, if you go to the Air Operations
Center or the Maritime Operations Center, the Tactical
Operations Centers and you see how we pass data, you are right.
We have got a long ways to go and we have to get there,
especially with the way our adversaries are moving.
You know, the decision advantage, there are two different
terms, Fully Networked Command, Control, and Communications,
FNC3, or JADC2, which is the Joint All-Domain Command and
Control system. That is about passing information. That is
connecting sensors of all types, from all varieties, from all
domains, from all services, and from allies to the right nodes
that can engage in the right nodes, it can do the command and
control. And that is the part we have not gotten to yet.
Senator Scott. Mr. Smith and Dr. Schmidt, would that be
true in your companies? Would you not be able to share data the
way the military has inability to share all information? And
something that is way more important than how well you run a
company.
Dr. Schmidt. Well, information is incredibly important. As
part of my DIB work, we spent a lot of time on this. Part of
the problem here is that the military has systems but does not
have software, and the systems have information and the
information has to go from one system to the other. So a series
of projects, they are generally known as Kessel Run and so
forth--they are well known to the staff here at the committee--
we are able, with relatively simply programming, to really,
really improve the lethality and the usefulness of these
systems.
Over and over again, the problem is that the military
thinks software is not valuable and it sort of collects it. I
propose that anybody who is in charge of a combatant command
(COCOM), in fact, any four-star general, should have 50
software programmers to just solve problems. And whenever that
has been done, the force productivity has risen very, very
quickly. So I used the TRANSCOM example before. It is a
relatively straightforward thing to have programmers write the
code to take to our enlisted people and have them do something
more useful than just copying numbers all day.
Mr. Smith. And I would add different categories of
information require different approaches. One of the concerns I
was raising before is when we think specifically about threat
intelligence, really the data about foreign cyberattacks on the
United States, the information is very much in a set of silos,
in the public sector and in the private sector. And I just
think it is actually worth pulling out the 9/11 Commission's
report, because I think it does speak to us, almost 20 years
later. What they said was that the government needed to move
from a culture where information was shared only when there was
a need to know to a culture of a need to share. And we have to
do it with privacy controls. We have got to have the right
division between the public and private sectors. But we are
only going to understand our threats better if we are doing a
better job of aggregating data and then harnessing things like
AI to alert us to what is happening.
Senator Scott. Thank you.
Chairman Reed. Thank you, Senator Scott. And now via WebEx,
Senator Rosen, please.
Senator Rosen. Thank you, Chairman Reed, Ranking Member
Inhofe, and, of course, all of the witnesses for being here
today. I really appreciate.
I really want to talk about international standards and
emerging technologies, because international standards, they
serve as the foundation for the development and the use of
emerging technologies. Our global competitiveness, it depends
on our participation and in our leadership in setting the
standards for the next generation of technologies. That is why
last year I helped introduce the bipartisan Promoting the
United States Wireless Leadership Act of 2020, to ensure that
U.S. has a seat at the table in the wireless standards-setting
process.
China has an explicit plan to become a standards-issuing
country by targeting emerging technologies, where global rules
have yet to be fully defined. For the U.S. to remain the leader
in this space, to maintain our national security edge, our
response must include working with the private sector,
investing in R&D and emerging technologies, coordinating with
relevant agencies, and engaging in international standards-
setting bodies. And as a former software developer I love the
comment that we should have 50 programmers embedded in all
these places. Programmers and analysts are key to solving so
many critical issues.
But my question is for Dr. Schmidt and then Mr. Smith.
Could you talk about the importance and the impact of U.S.
participation in the international standards-setting bodies for
the development and use of emerging technologies, and how
should we, as the government, be coordinating with the private
sector to really set those standards for the next generation
technologies?
Dr. Schmidt. Your diagnosis of the problem is exactly
right. It turns out that China now has a deliberate goal of
basically participating at a significant level at all of the
important standards-settings bodies, the most interesting being
5G Infrastructure Public Private Partnership (PPP), which is
the one that sets the 5G standards, where they now have figured
out a way to have a majority of the members. So that does not
bode well for the kind of values that we care about getting
embedded in these standards.
There are quite a few organizations, NTIA and others, that
are in charge of these, and I think that this is a good project
for the government to get itself organized around which are the
ones that are most important, because there are so many. Brad?
Mr. Smith. I would absolutely second that. First of all, I
think it is such an important question because it is easy to
overlook just how strategically important it is to the future
of American technology for the country to be successful in
influencing and helping to set international standards. It is
not a case of all technologies being equal, so as Eric
mentioned, you have to identify the technologies that we want
to prioritize. Different standards are set by different
standards-setting bodies, so then one needs to have an
engagement strategy. And certainly you need to think about how
to bring together the resources in the Federal Government in a
place like the National Telecommunications and Information
Administration (NTIA) and in the private sector, and we need to
do this by continuing to work with our allies especially.
The Chinese government has established for itself a
leadership role. It is going to use its own standards-setting
ability for its market to try to influence global standards,
and we need to be allied with our partners and working together
to ensure that we win the race to influence standards.
Senator Rosen. Thank you. I am going to build on that with
our STEM workforce shortfall, because in order for us to
continue to be the most innovative country, to set the
standards that we need to, we have to maintain a workforce that
can innovate. In the United States we are expected to face a
shortfall of nearly 3.5 million skilled technical workers. That
is just by next year. To address this shortfall, I introduced a
bipartisan bill called the PROMOTES Act, that is going to
authorize the Secretary of Defense to enhance the preparation
of Junior ROTC students for training and education in STEM
fields. I am proud that this bill was signed into law in last
year's NDAA, but more needs to be done if we are going to do
all the things we need to.
So, General Carlisle, can we talk for a moment about how
the Junior ROTC program, how we can leverage that to
incentivize, train our high school and college students to
enter these emerging technology fields like artificial
intelligence, quantum computing, cybersecurity, and so many
other spectrums? What role can the military play? How do we get
the workforce that we need?
General Carlisle. Thank you, Senator. I could not agree
more. I think our ability to attract the talent and bring them
into the STEM career fields, in particular. We, in the Air
Force, face--well, actually all services face a severe pilot
shortage, less so now, obviously, because most of the airlines
have not hired, but that will, I think, come back.
But one of the things is how do we get to those folks that
do not know about us. How do we get those communities that do
not have the opportunity and maybe do not understand what those
opportunities are in the military? Recruiting people, the
Junior ROTC program, a very good friend of mine runs the Air
Force ROTC program out of Maxwell Air Force Base, and what do
we do to attract these folks, to let them know there are
opportunities out there, and that the military can open up
training opportunities, it can open up different educational
opportunities, it can open up career fields to them that they
are not aware of.
So I think the military can play a huge part of that, and
as was mentioned earlier, I think it is K-12 is where it has to
start and then it goes to the world-class universities that we
have in this country and how they continue to attract, continue
to promote, and continue to be the leaders in their fields.
Again, I think the ability to get to the communities, because
we have, you know, the incredible population of this country,
and a lot of it is they just do not know. They do not know what
those opportunities are out there, and I think Junior Reserve
Officers Training Corps (ROTC) is a great way to start opening
up those opportunities.
We did start, for the flying piece, we started a program
with the Civil Air Patrol that would allow folks that could not
afford to go get a pilot's license, because it is not
inexpensive, at the cost of the program, go get a private
pilot's license over the summer and learn about aviation, and
then the ability to bring them back in to aeronautics or
astronautics or aviation is another opportunity for them that
they probably would not know existed beforehand.
So I think it is about making opportunities and getting to
the full breadth and width of the American population and offer
them those chances.
Senator Rosen. Well, thank you all. My time has expired but
I am excited to work on all of these issues with all of you.
Thank you, Mr. Chairman.
Chairman Reed. Thank you, Senator Rosen. Senator Hawley,
please.
Senator Hawley. Thank you, Mr. Chairman. Dr. Schmidt, let
me start with you. I am very concerned about the consolidation
of the defense industrial base. This is a multi-decade problem,
one that has really accelerated in recent years. And we are
seeing this problem now with emerging technologies, the subject
of this hearing today, where just a few large companies, like
the ones that, frankly, you represent, or have represented and
worked for, own a lot of the technology or can buy it up.
Two years ago, the Chairman of the Joint Chiefs and the
Secretary of Defense sat right where you gentlemen are sitting
and complained about Google, in particular. I was so struck
that I went and I pulled the transcript. The Secretary of
Defense said, ``I am talking about Google and their support to
China and their lack of support for the Department of
Defense.'' The Chairman of the Joint Chiefs, General Dunford,
said, ``The work that Google is doing in China is directly or
indirectly benefitting the Chinese military.'' Then he went on
to say, ``We are watching with great concern industry partners?
work in China, knowing that there is indirect benefit.'' And,
of course, Project Maven is what they were talking about the
time but there is also the controversy about Boston Dynamics
and the robotics collective.
Here is my question. How can we ensure robust competition
so that we have a competitive market for emerging technologies
that is not dominated by just a few big firms?
Dr. Schmidt. Well, first I am no longer at Google, and I
disagreed with the activities that you were describing, and
indeed I worked for the DOD during that period, so my personal
views are clear. I think there is good news----
Senator Hawley. Do you think Google made the wrong
decision--sorry, is that what you are saying, Dr. Schmidt?
Dr. Schmidt. Let me just leave my statement as what I said.
Senator Hawley. Well, I did not hear your statement here on
the record now, so just reintroduce it. Why do not you answer
my question? Are you saying that you disagree with----
Dr. Schmidt. I disagreed at the time with the decisions at
Google.
Senator Hawley. That the Chairman of the Joint Chiefs and
the Secretary of Defense were talking about, just to be clear?
Dr. Schmidt. Yes, that is correct.
Senator Hawley. Okay.
Dr. Schmidt. And it is important to know that during that
time I was an employee of the DOD, so my view is clear.
So with respect to--there is good news, that there are
plenty of companies that now want to work with and for the
military. Part of the problem they have is they are having
trouble getting through the valley of death. They have a good
idea. They cannot get into the right procurements. They do not
have access. The DOD has set up a set of initiatives, DIU being
one, and there are a number of other ones that are quite good.
And so I think to the degree you have a concern about
concentration around, for example, Google, your best strategy
is to have as many touchpoints where private sector innovators
can work with the DOD.
I should also note that Google's competitors, Microsoft and
Amazon, made very different decisions than Google did during
that time.
Senator Hawley. Let me ask you, Mr. Smith, speaking of
Microsoft, the use of Chinese-made hardware like printed
circuit boards, poses a significant cybersecurity concern for
the United States. I think some of my colleagues have mentioned
this earlier. Does Microsoft use Chinese printed circuit boards
in the systems you provide to the Department of Defense?
Mr. Smith. I would have to go look specifically. We have
been diversifying our----
Senator Hawley. Well, just before you move on from that,
will you do that and get me an answer on that question?
Mr. Smith. Sure. I would be happy to.
Senator Hawley. Great.
Mr. Smith. I will say, more broadly, two things are
important. One is we, like other companies that produce
hardware, have been diversifying our supply chain, which means
less reliance on China, more focus, including on printed
circuit boards, from Taiwan, as well as in other countries in
Southeast Asia and Mexico, and even we are looking at the
United States itself.
The second thing I would say is for anything that is going
to involve national security system, use for, say, the U.S.
Army, you know, every component is reviewed by the U.S.
Government itself in terms of where we are sourcing it.
Senator Hawley. I am glad to hear about your
diversification, and I heard your remarks on that earlier. Let
me just press you on this point, though. Will you commit to
ending Microsoft's use of Chinese printed circuit boards if, in
fact, you are still using them?
Mr. Smith. I would like to learn more. I would be happy to
send you a letter and we will give you a commitment. I believe
we may no longer be using any printed circuit boards from
China, but I would like to go look.
Senator Hawley. That would be good. That would be good. If
you are, though, will you commit to ending the practice?
Mr. Smith. I have learned enough over the years that I
should be informed by the other employees at our company before
I give a definitive answer, but I will be happy, Senator, to
give you a definitive answer.
Senator Hawley. Okay. You are not going to give me one here
today, though, it sounds like.
Mr. Smith. I would like to give you an informed and
definitive answer.
Senator Hawley. Uh-huh. Yeah. We hear that a lot before
this committee. Would you at least commit to being transparent
and notifying DOD about which systems contained Chinese printed
circuit boards, if, in fact, you are continuing to use them?
Would you give me that commitment?
Mr. Smith. I believe we already are. If we are not, that
is--of course we want to be transparent with Department of
Defense (DOD) with all of the components that are going into--
--
Senator Hawley. Okay, good. So yes, you will do that.
Mr. Smith. Yes, I will do that.
Senator Hawley. Okay. Outstanding.
Mr. Chairman, I see that my time has expired. I have got
some more questions for you, Mr. Smith, and also for you, Dr.
Schmidt, but I will give them to you for the record. Thank you
for being here and thanks for your work. Thank you, Mr.
Chairman.
Chairman Reed. Thank you, Senator Hawley. Senator Kelly,
please.
Senator Kelly. Thank you, Mr. Chairman, and thank you, Dr.
Schmidt and Mr. Smith and General Carlisle. And, Mr. Chairman,
I look forward to serving on this committee.
And, General Carlisle, in your opening testimony you
mentioned that we are lagging behind our adversaries in a
number of areas--hypersonics, directed energy weapons systems,
and microelectronics. About 18 months ago, the People's
Liberation Army (PLA) fielded what is perhaps the world's first
operational hypersonic weapon system, DF-17. Has a hypersonic
glide vehicle as well, and that vehicle can suppress its entry
trajectory and accelerate to Mach 5. Intercepting this vehicle
with existing anti-ballistic missile (ABM) technology is
incredibly challenging, and we do not currently have a defense
against that, as far as I know. It has a range of thousands of
miles, putting our assets and our troops and our equipment in
Japan and South Korea at great risk.
As a former commander of the Pacific Air Force, how big of
a strategic impact is this in the theater?
General Carlisle. Senator Kelly, it is a tremendous impact.
It is a tremendous impact to all the entire joint force and the
ability to operate. You have heard before us talk about the
ability of the adversary to deny us entry into the space,
whether it is by a naval--by air anti-satellite weapons is
another case where they deny our ability to use a domain via
laser or on orbit or direct descent at us, anti-satellite
weapons. So it was a huge impact, and clearly, as I mentioned
earlier, where China has come over the last 20 years in their
fielding of capability at a pace that is extraordinary, it has
changed the dynamic in the Pacific tremendously.
And the earlier question, I think one of the things that it
is incumbent upon all of us, and certainly this body and use
that have the opportunity to still work in the defense
industrial area, is we have to educate the American population
on what the Chinese are attempting to do, what they have
written they want to do, and what they are blatantly going
forward with, that is counter to our values, our way of life,
and our future. The DF-17, the ability to sense where they are,
what they are doing, and then defeat them is a tremendous
challenge, and sir, we will come back and at a classified level
we can talk at a different level of what it did. But, I mean,
when you think about our ability operate again via the maritime
domain or the air domain or the land domain, it significantly
impacted and changed the concept of operations for engagement
in the Pacific.
Senator Kelly. Later I would like to talk to you about how
do we catch up. You know, how do we build a system, a defensive
system, but also how do we match that capability, or exceed it.
General Carlisle. Sir, I would love to come over and talk
to you about it.
Senator Kelly. And I have a couple more minutes. I want to
follow up on Senator Hawley's question a little bit,
semiconductor technology. And the CHIPS Act appropriated--did
not appropriate--authorized about $10 billion to manufacture,
to bring that manufacturing capability to the United States.
The Taiwan semiconductor manufacturing company has a 5-
nanometer chip that they currently make. It is my understanding
that Intel and other companies cannot manufacture a 5-nanometer
chip.
Can you outline, Mr. Smith, for us just where--and Dr.
Schmidt as well--just what technologies, and what is the--and
we only have about a minute left--what impact does that have
for our country?
Mr. Smith. Well, I do think you are right to identify this.
It creates a weakness and a vulnerability for the country, and
I do think a critical issue for the next couple of years is
going to require decision-making on how to catch up in that
space. Part of it is an issue of innovation, as you identified,
the gap. But I think another part does involve investment, and,
you know, Microsoft is obviously not in this part of the
technology business, but if we are going to bring semiconductor
manufacturing back to our shores I do think it is going to
require some targeted Federal investments, and it is not going
to be inexpensive. The kinds of dollars you were just talking
about I think captures well just how enormous it is in terms of
cost to build these kinds of fabrication capabilities.
Senator Kelly. Dr. Schmidt?
Dr. Schmidt. The CHIPS Act is a very good first step but it
is not enough. The 5-nanometer technology at TSMC is the world
class. They are now working on 3-nanometer technology, which is
allegedly going to be available within 12 to 18 months.
I have often wondered why is it that one group can stay
ahead, and the answer is that is year after year of precision
and learning and proprietary innovation and so forth, and
something which is very hard. Remember that the Chinese had,
for 30 years, a goal of catching up to TSMC, and they have
required, for example, fabs in China and so forth and so on,
and they still have not been able to do so.
So I suggest that what we do is we take American ingenuity,
which is profound, with some form of incentive system to sort
of close this gap, and put those semiconductor operations, at
least foundries, in the United States, and use them for both
commercial but also military purposes. It is critical that our
military chips be made in the United States, for the reasons
that everyone here would fully understand.
Senator Kelly. Thank you.
Chairman Reed. Well, thank you, Senator Kelly, and thank
you also for sitting through the hearing. I think you got some
practice sitting for hours in a cockpit, which prepared you
well for this committee.
Senator Kelly. And alert.
Chairman Reed. Senator Tuberville, please.
Senator Tuberville. Thank you, Mr. Chairman. Good morning,
guys. I know it has been a long--very quickly, you know, your
testimony today, I just hope everybody is listening across the
nation. We are in trouble. Our country is in trouble, and it is
going to be solved a lot by our technology. Most of us in here
went through a little bit of Vietnam and all these wars, these
no-nonsense wars that we have had over the years, and we have
wasted a lot of money on these wars, and we have gotten behind
China. We have not spent enough money, because we have not had
it.
But thank you for being here today, and Dr. Schmidt, I
enjoyed listening to you. In my former life of coaching I
learned a long time ago it is not about the money, it is about
organization. And if you are not organized you can throw all
the money at it you want, but you are not going to survive. So
I really enjoyed hearing that.
You know, in Huntsville, we lead the nation in many
categories in technology, so if you have not had a chance to
visit, it is the Silicon Valley of the South, I invite you to
come.
So just a couple of questions. Mr. Smith, the phrase
``American ingenuity'' during my lifetime rose, and we all saw
it grow and prosper. We thrived in an environment with less
regulations, smaller government, risk-taking. Silicon Valley in
the '80s and '90s worked much the same way. How do we get that
back? How do we get that back to where we can continue to grow,
instead of just the big companies? We have gotten away from it,
of the smaller companies just being able to innovate and grow
with us technology-wise. Because we have got to catch up,
somehow, some way.
Mr. Smith. Well, I think we still live in a country that
rewards people with bold ambition and the determination to make
that kind of dream come true. And, you know, when I joined
Microsoft we had about 4,000 employees. This was 27 years ago.
Today we have 165,000. It is a much bigger place, to your point
about organizations.
Senator Tuberville. What a country, right? What a country.
Mr. Smith. Yeah. But, you know, there are days when I still
feel like it is the smaller place. I think that is American
ingenuity, that spirit of creativity. And one of the
interesting things about the tech sector is it is an ecosystem.
You know, Eric has talked about this for years. You cannot
succeed at a big company unless you work closely with a network
of small ones. And I think one of the interesting things about
the NDIA is it really is the voice, in so many ways, of the
small defense contractors.
I think we should not worry for the need for the government
to invest more in large companies, absent, say, things like
chip fabrication. What we should look at is where the
government can ensure that there is an opportunity for small
companies, and then I would say for everybody across the board,
so we can go to the great universities, the community colleges,
and basically hire the talent we need.
Senator Tuberville. I had the opportunity to travel all
over, and campaigning the last two years in Huntsville, going
to 800 or so defense contractors, and, of course, National
Aeronautics and Space Administration (NASA), SpaceX, Blue
Origin, all of those, and it is amazing the technology that we
have. But it is also amazing, you know, what the private sector
can do, just going through the new laser technology that you
are seeing now, that our soldiers are going to hopefully be
able to use in the very near future, and hyper-ballistic
missiles. You know, we are behind China. You know, the general
was saying we are the best equipped, but we are getting old.
Our equipment is getting very old, and we need to do a lot of
things with that.
Dr. Schmidt, you say Americans can compete and win on any
playing field, and I know a little bit about that. But we have
seen China that is willing to cheat to win. They are willing to
steal our technology, use our own capitalistic system against
us. But I know that there are no shortcuts in winning. So if
you want to win you have to put out the work. How do we work as
a team better? You know, my question is this country is best
when our teammates work together, and our allies work together.
Do you think we are doing that very well?
Dr. Schmidt. There are parts where we are and in many
places we are not. I would urge, collectively, that we identify
bipartisan agreement around the areas where we must win. We
have mentioned hypersonics multiple times. Frankly, we have to
win there. What is our strategy to win? How are we going to get
there? We cannot spend 15 years building the first hypersonic
weapon while China and Russia are already working on it. We
need a different methodology.
So necessity drives the urgency and urgency then drives the
outcome. There are plenty of ideas of how to do it. You can do
it in a private model in a secure facility. You can do it
through the government, what have you. But the urgency should
drive it. The 5G issue that I highlighted, the issue of AI
leadership. In our AI recommendation we speak about doubling
the R&D budget for AI, which these numbers are small relative
to the Federal budget, but it would be hugely leveraging. There
is a list.
But the bipartisan consensus should be to build a national
competitiveness approach, literally globally competitive, all
of our technologies to wins, the military benefits and our
industrial base wins as well.
Senator Tuberville. Thank you, gentlemen.
Chairman Reed. Thank you, Senator Tuberville. Gentlemen,
thank you for your extraordinary testimony. It has been
illuminating. You have provided us extraordinary insights, but
also you have given us a long to-do list. So we appreciate that
too, and we look forward to working with you as we approach all
these problems.
Thank you. I have got to depart, along with my colleagues,
to vote, but I appreciate very much your participation, and
again, this was an extraordinary hearing because of your
insights, all of you. Thank you very much.
The hearing is adjourned.
[Whereupon, at 11:48 a.m., the Committee was adjourned.]
[all]