China treats AI more as infrastructure than something to directly profit off of, which is why this is a bit of a strange article. Their models are open weight generally, you aren't legally allowed to replace a job with AI, and their models are optimized to be low-cost and efficient compared to western models. This isn't a recipe for profit, but for public utility. It's expected that The Guardian would frame their strategy from the viewpoint of capitalists seeking to maximize profits, though. It's the same as when the CPC deliberately sank the cost of housing, which was being used as an investment vehicle, and capitalists complained about the market rather than highlighting the new affordability of housing.

I still don't know why it's so hard for people to recognize that AI is incompatible with capitalism. You can't just have an advanced autonomous system that increases competition drastically for extremely cheap and accessible be something profitable. It's as if capitalism is forgetting that humans need to earn money to pay for things?

China AI ain't gonna win. American AI ain't gonna win. AI gonna win.

So AI is the modern equivalent to the US/USSR space race? Bankrupt your opponent into investing into a non-viable tech?

Is this AI thing a new Starwars campaign like the one that brought down the Soviet Union? If so, who’s to win? If anyone that is.

The issue is the promised profits (edit or lets say productivity to avoid capitalist terms) simply never materialized and its a heavy burden on an over leveraged nation like china. US can buoy massive losses by pumping USD but that can only last so long too. AI is here to stay but the costs to run it, and the value you get out of it, are deeply imbalanced right now and portfolios are starting to show it. Name a VC who thinks AI is hot, none. Anyone with data centre stocks likely feeling some red lately. Its been estimated AI needs to generate 6 trillion in the next few years to cover data centre losses alone https://futurism.com/future-society/data-centers-trillion-capital-time-bomb-bain meanwhile revenues are trending downward. This is very similar to the Dot Com bubble.

Edit: public utilities still have costs of operation even if not pursuing profit. More expensive operations still drag down an econony which is why if you build a state owned power plant you should aim for efficient ones not inefficient ones. These data centres are exacting mass cost on the chinese economy which is energy insecure. These pressures cant simply be ignored by saying oh its okay its a public utility. If we had infinite energy and resources sure. But no one does.

Is this just due to the stockpiling of chips and data centers?

I guess it makes sense coming from a country that built empty cities, but are they speculating the was the west is?

Like where were they hoping to make money?

What I've realized is that because of the hype, you've virtually always got investors willing to offer a service for free, in hopes of attracting users.
Which means, if that service then tries to make that money back, users will quickly find another free offering and switch to that instead.

Obviously, in theory, you might manage to offer such a unique service that folks won't find a competing service, but:

1. a lot of these AI services do the same shit, because it is just language models and image models at the end of the day, and
2. if you can quickly build that service then others can quickly build it, too, so if people are willing to spend money on your offering, you won't need to wait for competitors.

This is a very different situation than a gold rush, and yet it feels ever more appropriate when folks say only those who sell shovels earn money from it...