U.S. Announces It Lost Something It Never Had

The U.S. lead over China in AI is gone. Not shrinking. Not eroding. Gone.

This assumes we had a lead. Bold claim from people who can't get a chatbot to stop hallucinating legal citations or telling users to eat rocks. China's been training models on more data with more compute while American companies argued about whether AI should be allowed to say mean things on Twitter. Fantastic prioritization. Really inspired stuff.

The op-ed calls for a change in national strategy. Here's a strategy: maybe don't spend three years letting venture capitalists burn $80 billion on chatbots that summarize emails. Maybe don't let your entire AI policy get written by the same people who thought crypto was the future. Maybe don't treat semiconductor manufacturing like a hobby you'll get around to eventually.

China built chip fabs. We built content moderation policies.

They trained models on everything they could scrape. We had ethics panels debate whether training data violated the spirit of consent. They threw compute at the problem until it worked. We threw McKinsey consultants at PowerPoint decks about responsible innovation frameworks.

Now some guy wrote an op-ed saying we need a new strategy. Perfect timing. Really proactive. The patient's been flatlining for twenty minutes but at least we're discussing whether to try the defibrillator.

Retail traders will read this and buy more NVDA calls because they think AI means number go up. They don't know what a tensor is. They don't know what compute means. They just know Cathie Wood said something about exponential growth in 2021 and they've been underwater ever since.

The lead is gone. The strategy failed. The op-ed got written. And somewhere in Shenzhen, an engineer who's never heard of responsible AI frameworks just shipped another model Americans will pretend doesn't exist until it's running their infrastructure.

Photo by on Unsplash