1、C O R P O R AT I O N RAND WALTZMAN, LILLIAN ABLON, CHRISTIAN CURRIDEN, GAVIN S. HARTNETT, MAYNARD A. HOLLIDAY, LOGAN MA, BRIAN NICHIPORUK, ANDREW SCOBELL, DANIELLE C. TARRAF Maintaining the Competitive Advantage in Artificial Intelligence and Machine Learning Limited Print and Electronic Distributio
2、n Rights This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited. Permission is given to duplicate this document for personal use only
3、, as long as it is unaltered and complete. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial use. For information on reprint and linking permissions, please visit www.rand.org/pubs/permissions. The RAND Corporation is a research org
4、anization that develops solutions to public policy challenges to help make communities throughout the world safer and more secure, healthier and more prosperous. RAND is nonprofit, nonpartisan, and committed to the public interest. RANDs publications do not necessarily reflect the opinions of its re
5、search clients and sponsors. Support RAND Make a tax-deductible charitable contribution at www.rand.org/giving/contribute www.rand.org For more information on this publication, visit www.rand.org/t/RRA200-1 Library of Congress Cataloging-in-Publication Data is available for this publication. ISBN: 9
6、78-1-9774-0525-8 Published by the RAND Corporation, Santa Monica, Calif. Copyright 2020 RAND Corporation R is a registered trademark. iii Preface Artificial intelligence (AI) technologies hold the potential for becoming critical force multipliers in future military capabilities. Indeed, China has id
7、entified AI as key to enhancing national competitiveness and protecting national security and has put forth a national AI plan representing a whole-of-society approach that is backed by significant investments. The plan is guided by a set of milestone goals intended to position the Peoples Republic
8、of China as the worlds leader in AI innovation by 2030. In spring 2018, given the potential significance of AI technologies for the U.S. Department of Defense (DoD) and national security, the Vice Chief of the Air Force asked RAND to examine the competitive standing of the United States relative to
9、China. If DoD has a competitive advantage, what should it do to maintain it? If DoD does not have a competitive advantage, what should it do to achieve and maintain it? The results of this research will help inform the choices of the U.S. government in general and DoD in particular regarding ways ac
10、tivities, posture, and capability developmentthat AI and machine learning technologies can be advanced and countered to meet present and future security challenges. No expertise in AI is required to read and benefit from this report. The research reported here was sponsored by the Vice Chief of Staf
11、f of the Air Force and conducted within the Force Modernization and Employment Program of RAND Project AIR FORCE as part of a fiscal year 2018 project, Maintaining the Competitive Edge in Artificial Intelligence and Machine Learning. RAND Project AIR FORCE RAND Project AIR FORCE (PAF), a division of
12、 the RAND Corporation, is the U.S. Air Forces federally funded research and development center for studies and analyses. PAF provides the Air Force with independent analyses of policy alternatives affecting the development, employment, combat readiness, and support of current and future air, space,
13、and cyber forces. Research is conducted in four programs: Strategy and Doctrine; Force Modernization and Employment; Manpower, Personnel, and Training; and Resource Management. The research reported here was prepared under contract FA7014-16-D-1000. Additional information about PAF is available on o
14、ur website: http:/www.rand.org/paf/ This report documents work originally shared with the U.S. Air Force on January 28, 2019. The draft report, issued on December 31, 2018, was reviewed by formal peer reviewers and U.S. Air Force subject-matter experts. iv Contents Preface . iii Figures and Tables .
15、 v Summary . vi Acknowledgments . x Abbreviations . xi 1. Introduction . 1 Research Questions . 2 Methodology . 2 Outline of This Report . 3 2. Comparing U.S.-China Artificial Intelligence Ecosystems . 4 National Artificial Intelligence Strategies . 4 Cultural and Structural Factors . 7 Bottom Line:
16、 How Do They Compare? . 14 3. Recommendations . 19 Axes of Competition . 19 Manage Expectations . 21 Create an Engineering Pipeline Under DoD Control . 23 Create and Tailor Verification, Validation, Test, and Evaluation Techniques for Artificial Intelligence Technologies . 24 Create Development, Tes
17、t, and Evaluation Processes for New Operational Concepts that Employ AI Technologies . 27 4. Conclusions and Future Research . 30 Conclusions . 30 Future Research . 32 References . 34 v Figures and Tables Figures Figure 3.1. Top 10 Countries Having Collaborations with Peoples Liberation Army Scienti
18、sts as Measured by Number of Jointly Authored Peer-Reviewed Articles . 20 Figure 3.2. Peer-Reviewed Articles Coauthored by Peoples Liberation Army Scientists and Overseas Scientists . 20 Tables Table 3.1. Prior Growth of Cross-Border Investments in Artificial Intelligence, 20132017 . 21 Table 4.1. A
19、llocating USAF Effort Levels Across the Spectrum of AI Technology Development . 32 vi Summary Artificial intelligence (AI) technologies hold the potential to become critical force multipliers in future armed conflicts. Indeed, the Peoples Republic of China (PRC) has identified AI as key to “enhance
20、national competitiveness and protect national security,”1 and has put forth a national AI plan representing a whole-of-society approach that is backed by significant investments. The plan is guided by a set of milestone goals and is intended to position the PRC as the worlds primary center of AI inn
21、ovation by 2030. If the plan is successful, China will achieve a substantial military advantage over the United States and its allies, with significant negative strategic implications for the United States. With the 2017 National Security Strategy and 2018 National Defense Strategy identifying China
22、 as a great power competitor and given Beijings focus on AI technology, how much of a lead does the United States have, and what do the United States and the U.S. Air Force (USAF) need to do to maintain that lead?2 For the purposes of this research, AI has been interpreted as the use of machine lear
23、ning (ML) technologies to address a variety of application domains and problems, resulting in a multitude of capabilities, such as computer vision, natural-language processing, decision support, and command and control. ML is the field of computer science concerned with creating programs that “learn
24、” from data using a large and evolving set of techniques grounded in statistics and mathematical optimization. Against this backdrop, we identified the aspects of AI and ML that need whole-of- government attention to accelerate U.S. investment and implementation, as well as the investments and subse
25、quent policies that would support U.S. commercial-academic-government AI and ML growth and protection. Our starting point was an assessment of current Chinese and U.S. AI and ML strategies; investment levels; and structural, systemic, and implementation differences that affect the development and ad
26、option of AI in both nations. Transforming AI advances into military capabilities requires leveraging advances in fundamental research or commercial industry, transitioning them to the military, assessing their effectiveness and suitability, and updating existing operational concepts or developing n
27、ew ones to take advantage of the new capabilities. We thus assessed the potential for U.S.-China competition in AI and ML along five main dimensions: breakthrough fundamental research; advances in civilian industry (private sector, state owned, or state funded); development and 1 State Council of th
28、e Peoples Republic of China, A New Generation Artificial Intelligence Development Plan, trans. Graham Webster, Rogier Creemers, Paul Triolo, and Elsa Kania, July 20, 2017. 2 The White House, National Security Strategy of the United States of America, Washington, D.C., December 2017; James Mattis, Su
29、mmary of the 2018 National Defense Strategy of the United States of America: Sharpening the American Militarys Competitive Edge, Washington, D.C.: U.S. Department of Defense, 2018. vii engineering to transition AI to the military; advances in validation, verification, testing, and evaluation (VVT an
30、d operational concept development. Our assessment is that, as of early 2020, the United States has a modest lead over the PRC in AI technology development. This is largely because the United States has had a substantial advantage over China in the advanced semiconductor design and manufacturing sect
31、or; the U.S. semiconductor industry is currently more capable and more advanced than that of the Chinese. A strong semiconductor industry is an essential foundation for good, solid AI research. China is attempting to erode this edge through massive government investment in the Chinese semiconductor
32、industry.3 Also, the Chinese semiconductor industry has the additional advantage of proximity to the enormous Chinese market.4 This situation is further aggravated by the concurrent lack of a substantial U.S. industrial policy.5 That being said, semiconductor design and fabrication are uniquely diff
33、icult processes. At present, Chinese firms continue to depend on American designs and lag behind South Korean and Taiwanese manufacturing.6 Chinese attempts to find a shortcut to superiority by investing in newer types of chips and computing (as opposed to trying to unseat reigning American and alli
34、ed companies in central processing unit production) may produce some fruit, but American and European companies are also highly active in exploring new computing technologies.7 China has an advantage over the United States in the area of big data sets that are essential to the development of ML appl
35、ications. This is in part because the Chinese regime and the large Chinese tech companies (such as Alibaba) are able to harvest much more personal data from the Chinese populace than U.S. tech firms can gather from the American populace because of the lack of real privacy laws and protections in Chi
36、na. Moreover, China has demonstrated both the capability and will to hack overseas databases so that it can leverage additional quantities of data. Also, the Chinese population is about four times 3 For example, In the period since September 2014, numerous provinces and municipalities have establish
37、ed their own IC integrated circuit Funds, or received capital from the National IC Fund to establish other IC-related funds. Reports on the establishment of IC Funds in Hubei, Fujian, and Anhui provinces indicate the high degree of Chinese government involvement in establishing the funds in order to
38、 meet national strategic objectives. According to the SIA Semiconductor Industry Association, provincial and municipal IC funds have raised a staggering summore than $80 billion. (Office of the U.S. Trade Representative, Findings of the Investigation into Chinas Acts, Policies, and Practices Related
39、 to Technology Transfer, Intellectual Property, and Innovation Under Section 301 of the Trade Act of 1974, Washington, D.C.: Executive Office of the President, March 22, 2018, pp. 9394) 4 Tekla S. Perry, “U.S. Semiconductor Industry Veterans Keep Wary Eyes on China,” IEEE Spectrum, October 10, 2019. 5 Perry, 2019. 6 Jeffrey Ding, Deciphering Chinas AI Dream: The Context, Components, Capabilities, and Consequences of Chinas Strategy to Lead the World in AI, Oxford, U.K.: Future of
sgpjbg002
工作日 8:30 - 17:30
会员动态：
报告分类
新质生产力
ChatGPT
新能源汽车
大模型
Sora
机器人
微短剧
芯片
薪酬
白皮书
低空经济
数据要素
创新药
人工智能
AI产业
信息科技
互联网
消费经济
汽车交通
电商零售
传媒娱乐
医药健康
投资金融
能源环境
地产建筑
传统产业
英文报告
其它
扫码登录
手机快捷登录
账号登录
