A few weeks ago, the White House released a long awaited report entitled “Winning the Race: America’s AI Action Plan.”  The represents “a decisive pivot toward deregulation as the key driver for rapid innovation in pursuit of global AI leadership.” 

The reaction on both sides of the aisle has been quite positive and reflects a consensus among “technology executives, national security analysts, and U.S. officials” about the importance of AI.  Which is particularly surprising in 2025, at a time when Republicans and Democrats can’t seem to agree on whether the sky is blue.     

Even the Washington Post published an editorial endorsing the plan.  “The most important question for the United States regarding artificial intelligence right now is not how it will be used, or even how it will affect the economy and culture. It is whether the U.S. will maintain AI dominance. The race to own the technology of the future is a race we must not lose.”  Which is especially interesting given the Post’s anti-Trump bent and the fact that that this plan “noticeably steps back from the Biden administration’s attempts to identify and mitigate potential risks and pitfalls of [AI].”

The fundamental reason why AI is such a hot topic right now is that, as an opinion piece in Forbes put it, “AI has the potential to change the world as dramatically as the groundbreaking technologies that sparked previous industrial revolutions.”  Wait a minute.  As dramatically as the industrial revolution?  Isn’t all “this talk of the fourth industrial revolution just hot air from marketers talking up the share price of hugely powerful global corporations” the Forbes writer went on to ask.  Some of it is, he concluded, but the breakthroughs are real and are coming at a faster and faster pace.  Google CEO Sundar Pichai has repeatedly said that the impact of AI on humanity “will be even more transformational than fire.”

Many people first became aware of AI’s power with the release of ChatGPT in November 2022, a “large language model chatbot” that can carry on a conversation, write a thank you note to your grandmother, instantly research and compose a 20 page paper on King Ethelred the Unready, create a short story about how a frog developed its love of eating mosquitos, write computer code, and much more. 

If you haven’t already experimented with ChatGPT or one of its many competitors, you should give it a try, right now.  Put aside this blog for a few minutes, download the free version and prepare to be at first amazed and then a little scared at what the program can do.  And each new version keeps getting better.  (ChatGPT5 was released last Thursday, August 7.)

When I first wrote about ChatGPT and the risks of AI in this blog in July 2023, I asked the question “Should you worry about AI?” and answered myself “Unless your job is threatened, my answer is not at all.”  That was then, this is now.  You should worry.  In the last two years AI has become so much cheaper and more powerful that even the Pope is worried. 

And China is already having a major impact on how AI develops and how it is used.  The country’s “Next Generation Artificial Intelligence Development Plan” was announced in 2017 with the same goal as Trump’s plan: to dominate worldwide AI.  One way they are trying to do that is by competing on price, as the Chinese company DeepSeek did in January when it announced a new product that was so cheap it rattled the stock market.  In fact, the Deep Seek announcement “triggered a $1 trillion sell-off in U.S. and European technology stocks in a single day… [because its cost and popularity] have raised concerns about the future market share of established tech giants, leading to substantial declines in their stock valuations.” 

Since then, “Chinese companies have flooded the market with 1,509 large language models… often using open-source strategies to undercut Western competitors.”  The most dramatic so far was two weeks ago (July 28), the release of GLM-4.5, by the Chinese startup Z.ai.  It makes “DeepSeek look expensive [with]… an 87% discount on … long conversations with AI.” 

One of the most frightening things about all these new Chinese products is that since 2022, the US has been trying to slow Chinese progress by restricting the sale of advanced AI chips to China.  But the onslaught of new products shows that “export controls clearly aren’t slowing Chinese AI development. A blacklisted company [Z.ai] just delivered competitive performance while operating under restrictions designed to cripple such capabilities.”

Does this mean China is winning the AI race?  Foreign Affairs recently published several articles addressing this question in a far more nuanced way than the title of this post or any newspaper headline.  In “The Real AI Race” Rand experts Colin H. Kahl and Jim Mitre take the discussion back to basics by reminding readers that “determining who is ahead depends on what it means to win.”  They note that these days most experts define the race based on when AGI (artificial general intelligence) programs prove to be at least as smart as human experts. 

Unfortunately, however, “there is no standard, shared definition of AGI or consensus on whether, when, or how [AGI] might emerge.”  Nevertheless, most experts have long been bullish on how quickly it will appear.  When I first wrote about AI in this blog in 2022, I quoted AI pioneer Herbert A. Simon’s 1965 prediction that that “machines will be capable… of doing any work a man can do… within twenty years.”  He was wrong.

In 2023, when the consulting firm AIMultiple reviewed 15 surveys completed by a total of 8,590 AI experts about when they think AGI will actually exist, the most frequent answer was remarkably similar to Simon’s –  about 20 years.   More specifically, they found that 50% of AI experts predicted “that AGI will probably… emerge between 2040 and 2050 and is very likely (90% chance) to appear by 2075.”  That leaves 10% who think AGI emergence will take more than 50 years, if it ever occurs at all. 

You can put me in with the skeptical 10%.  Perhaps I was scarred by my own brief experience in AI in the early 80s, when I worked for a consulting firm on an ill-advised AI project for the US Army.  Or my skepticism could be grounded in my firm belief that we simply don’t know enough about the human brain to expect to improve on it.  When I earned a Ph.D. in Psychology, I spent countless hours reading about the major breakthroughs in neuroscience that were just around the corner.  Today, 50 years later, as far as I can tell they are still around the corner.  Some have described the brain as the most complex structure in the universe, with over 80 billion neurons, many connected to thousands of others at 100 trillion synapses.  It’s a miracle we can walk and chew gum at the same time.

Not only is the race to AGI uncertain to ever end, but it oversimplifies the problem.  As Kahl and Mitre concluded “The notion of a singular AI race between the United States and China fails to capture the true complexity of the rivalry unfolding today. The challenge is to win not one definitive contest but a multifront competition whose outcome will shape the international balance of power.”  Their article reviews multiple AI races, some of which may be won by the US and others by China. “Militaries and intelligence agencies must harness AI’s transformative potential and mitigate its disruptive effects. Similarly, countries stand to gain a competitive edge if they can adopt AI at scale across the economy and society. Governments are also battling to create and own the standards, supply chains, and infrastructure that will undergird the global technological ecosystem.”

Trump’s plan for American AI dominance is based on the US tech industry’s approach of developing proprietary AI models.  This is certainly the road to being able to charge the highest prices.  For example, Apple’s proprietary approach to iPhones and other products has enabled them to charge premium prices and become extraordinarily profitable.  But there also huge risks to the proprietary approach.  If you are old enough to remember the 1970s fight over standards in videotapes, you may have owned a Betamax recorder, which was Sony’s attempt to establish standards for this fledgling marketplace.  While it was widely regarded as technically superior to the competing VHS videotape format, Betamax lost out for several reasons including its higher cost. 

Given multiple AI races, does it make sense to try to choose a single winner?  Brad Smith, the president of Microsoft, thinks so: “The number one factor that will define whether the United States or China wins this race is whose technology is most broadly adopted in the rest of the world.”  Eric Schmidt, former CEO of Google, agrees and goes on to argue that “Chinese AI firms have expanded their influence by freely distributing their models for the public to use, download and modify, which makes them more accessible to researchers and developers around the world.”  And, Schmidt says, China “has developed a real edge in how it disseminates, commercializes and manufactures tech. History has shown us that those who adopt and diffuse a technology the fastest win.”

Trying to win the AI race by addicting the world to proprietary standards defined by US firms, the way Trump’s plan does, may be aiming in the wrong direction.  According to still another Foreign Affairs article published recently, “America should aim for victory but prepare to finish second… Washington cannot and should not expect its lead to last forever… Washington needs to plan for a possible future in which the United States loses the AI competition to China—or, at the very least, one in which Chinese AI models are as popular globally… Finishing second is not a death knell for American AI, but refusing to adapt to compete would be.”