How America Can Beat China in the AI Race

 

If you believe Kai-Fu Lee — the Beijing-based AI scientist, venture capitalist, and former Googler (he ran Google China) — China has the edge in AI, a case he makes in “AI Superpowers: China, Silicon Valley, and the New World Order.” As Lee sees things, China’s internet economy generates lots of data with few restrictions on hoovering it up. China is also massively investing in the sector. In addition, he praises Chinese-style cutthroat entrepreneurship vs. the softer Silicon Valley version, at least as he see it. (This interview with McKinsey gives a good sense of Lee’s views.)

Whether or not Lee is correct in evaluating how the two nations stand in relation to each other, it makes sense for US policymakers to assume more could be done to reduce barriers to AI advancement and encourage the technology’s development. And to that, it’s certainly worth taking a look at “Reducing Entry Barriers in the Development and Application of  AI” by Caleb Watney of the R Street Institute. The analysis finds that one goal of policy should be to increase the supply of AI talent and speed AI diffusion by importing more AI talent and allowing companies to better deduct the cost of training AI workers.

On the data side, policymakers encourage the creation of open datasets and data sharing, both in government and in the private sector where regulation currently inhibits sharing. It’s also important to “maintain a healthy ecosystem around distributed platforms” such as Amazon Web Services and Google Cloud. Government must “be careful to avoid data-localization laws [and] excessive privacy laws,” as well as antitrust efforts against those tech companies leading the way in AI research and innovation.

For his part, Lee also thinks immigration is key to America’s AI performance, as well as greater government funding of AI research. As to Lee’s forecast, a cautionary note from The Economist:

True, AI represents the new space race, and China and America are set to lead it. But Mr Lee’s comparative analysis of Chinese and Western capitalism suffers (ironically) from a lack of data. China has had only 30 years’ experience of capitalism since Deng Xiaoping’s reforms took hold: not enough to discover whether its no-holds-barred approach is indeed more efficient than a rules-based system of competition. It took several nasty financial disasters in the late 19th and early 20th centuries for the West to constrain its own worst business practices, the better to let its animal spirits flourish. A financial crash may yet temper Chinese hubris. Despite its futuristic theme, Mr Lee’s book fits into a familiar genre of business scare stories. In the 1960s the French were aflutter about “Le Défi Américain” by Jean-Jacques Servan-Schreiber; in the 1980s Americans were paralysed by Ezra Vogel’s “Japan as Number One”. “AI Superpowers” should be taken seriously. But it is not the final word.

Published in Economics, Technology
Like this post? Want to comment? Join Ricochet’s community of conservatives and be part of the conversation. Join Ricochet for Free.

There are 8 comments.

Become a member to join the conversation. Or sign in if you're already a member.
  1. Guruforhire Inactive
    Guruforhire
    @Guruforhire

    Purloined Nazi scientists?

    • #1
  2. Hank Rhody, Red Hunter Contributor
    Hank Rhody, Red Hunter
    @HankRhody

    James Pethokoukis: The analysis finds that one goal of policy should be to increase the supply of AI talent and speed AI diffusion by importing more AI talent and allowing companies to better deduct the cost of training AI workers.

    I think you got a buzzword stuck in your system. A good Find/Replace will set you to rights in no time.

    • #2
  3. iWe Coolidge
    iWe
    @iWe

    Government Funding is a terrible solution. The last time it really worked to advance new technology was the Manhattan Project.

    The Chinese are absolutely terrible at new technology of their own. They are no threat in the AI sphere. None.

    • #3
  4. Percival Thatcher
    Percival
    @Percival

    AI has been the Next Big Thing for over 30 years.

    • #4
  5. Black Prince Inactive
    Black Prince
    @BlackPrince

    iWe (View Comment):

    Government Funding is a terrible solution. The last time it really worked to advance new technology was the Manhattan Project.

    The Chinese are absolutely terrible at new technology of their own. They are no threat in the AI sphere. None.

    @iwe How did you reach this conclusion?

    • #5
  6. I Walton Member
    I Walton
    @IWalton

    They’re lots, I mean really lots,  of really smart Chinese and many are educated here, but ultimately the cost of labor is so low there I should think it would  limit rational and dense use of it in China as labor saving.  Moreover it’s top down with state guidance everywhere.  With time that will return them to the Chinese disease– in which  extreme materialism and lack of countervailing values leads to deep uncorrectable corruption.   We must avoid doing the same by thinking the Federal government can help it along by making key decisions for the industry.  Just remove barriers to entry everywhere, and improve secondary education.  The issue of Chinese stealing and buying technology has to be addressed as well.   Their  export economy drives a lot of their strength and we can deal with that in ways that help without picking and choosing goods to restrict or countries to discriminate against, but we don’t even talk about such policies.

    • #6
  7. iWe Coolidge
    iWe
    @iWe

    Black Prince (View Comment):

    iWe (View Comment):

    Government Funding is a terrible solution. The last time it really worked to advance new technology was the Manhattan Project.

    The Chinese are absolutely terrible at new technology of their own. They are no threat in the AI sphere. None.

    @iwe How did you reach this conclusion?

    Observation.

    In my world, aerospace, Chinese have no innovation whatsoever. They steal and copy – badly. The culture seems to be thoroughly incompatible with invention: invention is highly dependent on individuals who are willing to see things differently from others, and bet on those differences.

    AI requires step-change thinking. I have no data to suggest the Chinese are capable of it. Step-changes in technology are not a function of dollars and manpower.

    • #7
  8. Black Prince Inactive
    Black Prince
    @BlackPrince

    iWe (View Comment):

    Black Prince (View Comment):

    iWe (View Comment):

    Government Funding is a terrible solution. The last time it really worked to advance new technology was the Manhattan Project.

    The Chinese are absolutely terrible at new technology of their own. They are no threat in the AI sphere. None.

    @iwe How did you reach this conclusion?

    Observation.

    In my world, aerospace, Chinese have no innovation whatsoever. They steal and copy – badly. The culture seems to be thoroughly incompatible with invention: invention is highly dependent on individuals who are willing to see things differently from others, and bet on those differences.

    AI requires step-change thinking. I have no data to suggest the Chinese are capable of it. Step-changes in technology are not a function of dollars and manpower.

    So how do you account for this? Is the Chinese mind in some way inferior to other minds? I’m not making a hidden accusation of racism here — my only goal is truth and understanding. From a security point-of-view, even if your observation is correct (I don’t believe it is, or always will be), it does not mean that America will ultimately prevail in our cold war with China — we have our own weaknesses that can be exploited. Furthermore, given America’s openness and economic integration with China, I assume that whatever we know, they know.

    • #8
Become a member to join the conversation. Or sign in if you're already a member.