AI, China & Copium

When it comes to technology in general, and artificial intelligence in particular, it has been surreal to watch events unfold. From politicians in technology backwaters to opportunists and fashion billionaires, everyone has jumped on the “AI” version of Shinkansen. And if that wasn’t enough, the arrival of DeepSeek has unleashed a wave of fear, nationalism, and old-fashioned protectionism.

DeepSeek makes two major points.

First, as noted earlier, DeepSeek represents more than just a product — it embodies a different way of thinking that challenges the conventional approach to AI. It’s not about a specific company, but about demonstrating that others can also leapfrog the big boys with clever thinking.

Second, for the past ten years, the Chinese have been relentlessly focused on their “Made in China 2025” initiative. It shouldn’t come as a surprise that the technology gap between the U.S. and China has narrowed. “China 2025 sets specific targets: by 2025, China aims to achieve 70 percent self-sufficiency in high-tech industries, and by 2049—the hundredth anniversary of the People’s Republic of China—it seeks a dominant position in global markets,” noted the Council on Foreign Relations.

If a relatively unknown Chinese hedge fund can achieve what it did with DeepSeek, imagine what the larger Chinese technology giants — Baidu, Huawei, Alibaba, and ByteDance — can do with AI. Artificial intelligence will likely lead to large-scale automation. Given China’s reliance on industrial output, it’s a safe bet that they are developing AI variations for robots, industrial machines, driverless cars, and large-scale automation.

Meanwhile, on the Western front, Sam Altman and Elon Musk remain locked in their battle of egos. Instead of creating best-in-class products that consumers crave, Google’s AI chief (and Nobel Prize winner) Demis Hassabis is busy touting his company’s cost advantage. Call me old-fashioned, but all this amounts to defensive thinking. The competition isn’t going away. The world, to many’s surprise, is changing.


It might feel a bit old-school blogging, the following articles and selected quotes serve to reinforce the points made in the preceding brief piece.


Bill Gates on AI:

It’s unbelievable how good humans are at discarding unimportant information and retaining what’s important. We evolved to find food and not get lost, things like that. And yet, if you’re in a meeting with four or five people, even 20 minutes later, if you say, “Okay, what was everybody wearing?” you probably won’t remember that. Whereas for the computer, because it has essentially infinite storage, it’s got an eidetic type memory where it’s all there. Humans are so good at pruning the memory down.

There are certainly a lot of innovative things that the world needs, whether it’s in diseases or clean climate technologies. But the most profound technology is AI, because it’s going to have such an impact on all the scientific work we do and even help do some of the jobs that we have to have humans do today. So I definitely recommend understanding where it works and where it doesn’t work, and there are going to be some great advances made in this next decade that make it, in some ways, even more important than the revolutions I got to be part of the digital revolution, the PC, the mobile phone, the internet. Now, of course, AI builds on all those, but in a sense it’s even more pro- found because it’s creating nonhuman intelligence.

It’s very difficult to predict because we’ve never had anything like this. I mean, people say that past advances have just broadened the job market, and people still work pretty long hours. But AI is kind of unbounded in terms of its actual capability.

Three Observations – Sam Altman:

The cost to use a given level of AI falls about 10x every 12 months, and lower prices lead to much more use. You can see this in the token cost from GPT-4 in early 2023 to GPT-4o in mid-2024, where the price per token dropped about 150x in that time period. Moore’s law changed the world at 2x every 18 months; this is unbelievably stronger.

Google AI Chief Says DeepSeek’s Cost Claims Are ‘Exaggerated’ 

“We don’t see any new silver bullet technologies. DeepSeek is not an outlier on the efficiency curve.”

Silicon Valley’s delusion machine:

Thanks to apps like TikTok, Shein, Temu, and, most recently, DeepSeek, we know that China has caught up to the US and its tech industry has figured out how to innovate in ways ours can’t or won’t. You might not like machine-learning-based short-form video apps or gamified social shopping platforms, but they are genuinely new ways of interacting with the web. And US regulators can’t actually stop the tide from turning — at best, the US will become an island surrounded by a global internet run by Chinese software. But what elevates this from lame to genuinely dangerous is that this delusion that Silicon Valley can now decide how the future should look has infiltrated the highest levels of the US government. And AI is the technology powering this delusion.

When Tech Bans Stimulate More Than Stymie China:

This paradox of Western efforts to stymie China’s tech development actually stimulating it is not new. As Zheng Yongnian, a Shenzhen-based scholar closely listened to by the Chinese leadership wrote in Noema in 2018, the tech restrictions already then imposed may slow the pace of China’s leap forward but will “not be able to stem its high-tech catch-up. China’s enormous reserves of state capital, its substantial pool of ready talent and its huge market will continue to drive it forward.”

DeepSeek signals the dawning era of APAC:

American apps, American companies, American consumers and American culture have dominated the global market for decades. Yet in recent years, we have seen shifts. Local-language TV shows on platforms like Netflix have disrupted the primacy of Hollywood. Music scenes have localised, with majors like Universal and Warner investing in independent labels in emerging markets to keep abreast of the shift. While the US remains the highest ARPU market for social platforms, Asia Pacific (APAC) is rapidly becoming the largest in revenues.

February 11, 2025. San Francisco

7 thoughts on this post

  1. “the US will become an island surrounded by a global internet run by Chinese software.”
    The above comment accurately reflects trends across various industries, including automotive, renewable energy, and solar. BYD cars are seen worldwide, except in the U.S.

    1. I don’t think that is is a problem. The problem is that we are not exposed enough to the rest of the world’s work and assume we are the best. This will be the challenge. I have taken a ride in a BYD EV. For the price, they run circles around Tesla and most definitely around the Germans.

  2. I think it’s broken into two pieces. basic AI tools as a commodity, and then the more nuanced approach of designing effective focused systems that can make use of it.

    Its that second part that is intriguing because too much of the coverage is focused on general tools, not single issue specific tools (like understanding an x-ray) that will have a better return on investment.

    1. Paul,

      What you say is 100 percent correct. Sadly honest, logical thinking around “automation” argument doesn’t get you valued at insane valuations. Sadly, these days US tech starts with valuations and not value.

    1. Amen to that Dana. Also, even if it is a big team — China has no shortage of engineering talent — it is still taking a different approach. I am not dismissing any of the charges levied against them. The reality is that they have shed the light on a different way of thinking — to me that is more interesting.

Comments are closed.