share_log

美国AI国家安全委员会:解密中美AI战

US AI National Security Council: Deciphering the Sino-US AI War

钛媒体 ·  Nov 18, 2019 23:59  · 热门

Author: Springwillow Wang Ji

Source: titanium media

The United States established the National Security Council for artificial Intelligence in 2018. Why should it focus on China, its biggest competitor? In the words of former Deputy Secretary of Defense Bob, "this is a game we may lose, and we really need to take it seriously."

In recent years, China has made rapid progress in research and development, investment and even industrial applications in the field of artificial intelligence, and has made remarkable achievements all over the world.

Spare no effort in investment, take 2017 as an exampleThe total volume of global artificial intelligence investment and financing has reached US $39.5 billion, of which the total investment and financing in China has reached US $27.71 billion. Chinese artificial intelligence enterprises account for 70% of the total global financing. 31% of the financing transactions took place in China.

From the perspective of patent application, according to the statistics of patent analysis institutions on the number of AI-related patent applications in the world during the three years from 2016 to 2018, China accounted for 19 seats in the top 50 private enterprises and the United States accounted for 12 seats. Compared with the last statistical cycle (2013-2015), the situation has been greatly reversed. The number of Chinese enterprises and patents far exceed that of the United States, in the high-tech field of artificial intelligence. China's sense of existence is getting higher and higher.

In China, artificial intelligence has become a national strategy in 2017, and in July 2017, the State Council issued a "New Generation artificial Intelligence Development Plan" and made it clear that it will seize the global commanding heights of artificial intelligence by 2030.

China has made national efforts to strengthen the research and development of artificial intelligence, which has attracted the attention of the US science and technology community and even the Trump administration.

The most representative is that in 2018, the United States set up a special committee, the National Security Council on artificial Intelligence (NSCAI) (hereinafter referred to as "the Committee"), to assess the level of competition in artificial intelligence in the United States, and to make recommendations to Congress to formulate a comprehensive global competition strategy for artificial intelligence.

The committee now has 15 members, chaired by Eric Schmidt, Alphabet Inc-CL C's former chief executive, and includes Bob Work, a former US deputy secretary of defense, as well as a number of tech industry leaders.

Titanium media found that in the year since its establishment, the committee has acted frequently according to public data. First, we have focused on the development of artificial intelligence since March 2019, and three plenary meetings have been held so far.

At the end of July this year, the committee submitted its first report to Congress, calling China the "most important strategic competitor" of the United States in the field of artificial intelligence.In early November, the committee issued another report urging the US government to "increase investment in artificial intelligence".More importantly, the committee recommended to Congress that "global leadership in artificial intelligence technology is a national security priority for the United States."

You may have the question: what is the composition of the committee? How will they influence the decisions of Congress and Trump? What strategy will the United States adopt to join the competition in the field of artificial intelligence in the future?

Recently, Craig Smith (hereinafter referred to as "Craig"), the US-based contributing author of Titanium Media, exclusively interviewed Eric Schmidt (hereinafter referred to as "Eric") and former US Deputy Secretary of Defense Bob Walker (hereinafter referred to as "Bob"), and exclusively authorized Titanium Media to release the Chinese version of the content. The two committee members disclosed to the outside world the core mission and the latest progress of the "committee".

What exactly are the responsibilities of the NSCAI committee?

The background of the establishment of the NSCAI Committee in 2018 is closely related to China. In 2017, the Chinese government announced that it would "become a leader in the field of artificial intelligence by 2030", which directly stimulated the determination of the United States to study competitive countermeasures.

Eric Schmidt made it clear in the dialogue that NSCAI is "studying issues related to artificial intelligence and national security on behalf of Congress." The specific tasks of the Committee are divided into two phases:

In the first stage, the task is to investigate and collect evidence. By studying different reports, we can find out "what is the level of artificial intelligence technology in the United States?"

The task in the second phase is to make "policy recommendations" to Congress.

Eric revealed that the first phase has come to an end. At present, the work of the Committee has entered the second stage. When studying the actual level of artificial intelligenceThe committee is currently working with the five major allies--Britain, the European Union, Japan, Canada and Australia have conducted relevant exchanges-none of which China for the time being.

Bob, a former deputy secretary of defense and a member of the committee, said in an interview that they have a clear priority-to maintain the advantage of US artificial intelligence technology in the field of defense. "the impact of artificial intelligence on economic prosperity also involves national security."

Draw a focus: the committee is "focusing" on national security issues.

Eric admits that the United States does need a collective strategy in the field of artificial intelligence, but it is not the direct responsibility of the committee to develop a strategy."I personally don't think [the committee] is going to come up with a national strategy for artificial intelligence in the United States," Bob said.Our position is to advise Congress and assist Congress in policy formulation and legislation to ensure as much as possible the competitive position of the United States in the field of artificial intelligence.

One year and eight months after its establishment, in November 2019, the Committee publicly released an interim report. The report publishes five directions of policy recommendations from the seven "consensus" and "next phase" committees, as well as 27 preliminary conclusions. Eric said that everything in the report is the consensus reached by the committee.

wm

The cover of the interim report released by the National Security Council on artificial Intelligence (NSCAI) in November 2019 (source: NSCAI website)

Can we continue to cooperate with China, the "biggest rival"?

Both Bob and Eric believe that China has an advantage in artificial intelligence technology.

Bob said that China's purchasing power parity (PPP), the equivalent coefficient between currencies calculated based on different price levels of each country, has exceeded that of the United States. If this trend continues, it is possible that the absolute value of China's GDP will exceed that of the United States by the end of the 1920s.

"the United States may face a competitor whose economic strength is stronger than ours. In addition to artificial intelligence, China hopes to become a world leader in innovation. " The former Defense Department official said.

Why does the committee spend so much time studying this "biggest competitor"?

Bob explained that"this is a game we may lose and we really need to take it seriously and try to stay in the lead.The common goal of the committee and Congress is to ensure that the United States has a "winning plan" in an important area of technology.

On the issue of how to cooperate with China in artificial intelligence technology, Bob said, "there are several channels of dialogue, but they do not include the committee." The committee has no plans for formal contacts and exchanges with China. "

One thing is clear:Both Bob and Eric are opposed to the complete suspension of cooperation and exchanges (technology, talent, etc.) between the US Congress and China.

Chinese talents, yes or no?

In terms of whether to continue to introduce Chinese talents and researchers, it is not difficult to see the ambivalence between American academic and political circles.

In November this year, US Ambassador to China Terry Branstad published a signed article "the United States welcomes Chinese students" in the China Youth Daily, which, of course, does not help ease the "tightening of visas for studying in the United States."

As soon as the ambassador to China spoke, NSCAI mentioned the issue of talent introduction in its latest report released in November. Eric Schmidt told the media that "the United States relies on Chinese technical experts, and decoupling at the talent level will hurt the United States."

Chinese students have always contributed a large number of talents to American high-tech research. Reuters also commented in a November report that there are deep talent, corporate and technological ties between China and the United States, and that once these ties are severed, American researchers and companies "may lose valuable markets, important talent pools, and ongoing research in Chinese laboratories."

In a conversation with titanium media contributing author Craig, Bob said that it is inevitable to face challenges in the development of artificial intelligence. "in addition to the threat posed by intelligence activities, we also face the challenge of the theft of artificial intelligence technology and China's artificial intelligence strategy." Under the above circumstances, it is certain to protect the national interests of the United States, "but cooperation can also bring many benefits, especially in the field of artificial intelligence."

Eric thinksIt would be extremely expensive to stop communicating with China completely.

Eric's argument is that Chinese talent is an indispensable resource for research programs in American universities. "even if there are things you don't think can be trusted, you can build a credible system." This is not a binary option that is either fully accepted or completely resisted. "

Eric believes that there can be a "middle ground" on the issue of cooperation or non-cooperation."for the United States, this is probably the best choice.Because our job is to make sure that the United States is the winner.Eric said.

Both Bob and Eric expressed the hope that the United States would continue to attract innovative talents from all over the world and maintain its appeal in the field of artificial intelligence.

Unfortunately, the NSCAI committee has not yet made specific policy recommendations to Congress on the introduction of talent.

Compared with China's "150 billion"Will the US government invest more?

"there is no specific number." That's what Bob told Craig, the titanium media contributing author.

According to China's artificial intelligence development strategy, it will invest a total of US $150 billion by 2030 before it becomes a global leader. Craig asked Eric and Bob this question of global concern:How much is the expected investment of the US government?Will it be close to China?

Obviously, Eric and Bob can't give a definite answer.

The two interviewees admitted that government spending in the field of artificial intelligence in the United States is already very urgent, and in fact, the committee has also studied the issue. "[the development of artificial intelligence] must spend money. Global leadership and artificial intelligence are national security priorities.

Bob tried to get a clear amount from Congress on spending, but "no government functional group specializes in this area."Eric said only that at this stage, the United States has not yet defined in government spending "within the scope of artificial intelligence planning". The committee is currently trying to work with a number of data science companies "to work together on related issues".

Eric admitted that Congress has received some relevant legislative proposals, such as "an additional $10 billion in annual spending" and so on. However, Eric is slightly evasive on the issue, giving Craig the questioner an example: in the case of the National Science Foundation, what is the amount of government spending? The foundation "doesn't even reach 10 billion dollars a year," and Eric seems to be hinting at a lack of confidence.

Eric said it was "necessary" to urge Congress to establish an artificial intelligence-related foundation, but "there is no such institution yet."

In fact, according to titanium media, in September this year, the US federal government calculated and announced for the first time specific requirements for AI spending by various agencies. "in the fiscal year up to September 2020, US government agencies requested to spend $973.5 million on non-defense AI research."AI spending involving the US Department of Defense is classified.

In September this year, the US federal government calculated the specific requirements of various agencies for AI spending, "requiring $973.5 million on non-defense AI by September 2020". However, the US Department of Defense's AI spending is classified (figure / Visual China).

Should AI Innovation be "far away from the Government"?How should the government make rational use of artificial intelligence?

With more and more application scenarios of artificial intelligence technology, data acquisition and privacy protection have surfaced, which has become a problem that governments all over the world have to face.

How can the government remain "transparent" in the application of artificial intelligence in various industries?

Bob cited the incident of Alphabet Inc-CL C's withdrawal from Project Maven (the "Project Maven incident" can be seen as a note in the dialogue below). The problem of data abuse and public trust has been widely seen as a dangerous omen.

After the Alphabet Inc-CL C incident, more private entities may choose to "stay away from the government" in their artificial intelligence innovation activities."fortunately, this has never happened." Bob said.

Before the establishment of NSCAI, the US Department of Defense also established the "National Defense Innovation Commission", which focused on formulating basic principles for the use of AI technology in national security affairs, and was responsible for communicating with the public to explain how they would "morally use artificial intelligence technology."

The members of the National Defense Innovation Commission and the National Security Council of artificial Intelligence overlap. As a result, NSCAI is also responsible for submitting "recommendations on the principles for the use of artificial intelligence" to the Department of Defense.

Judging from the progress since the establishment of the American artificial Intelligence National Security Council (NSCAI), the United States has made a response to the global artificial intelligence science and technology war, and an AI war between China and the United States has begun.

Edit / emily

The translation is provided by third-party software.


The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.
    Write a comment