American AI if the boom is a bubble: the Karp-Zitron scenario

According to LessWrong, the booming artificial intelligence sector may be built on fundamentally unsustainable economics. Two recent media appearances highlight the problem. Palantir CEO Alex Karp criticized the status quo: OpenAI and Anthropic host models for corporate customers, which lets those companies see their clients' business operations—and clone any successful use case they spot into their own products. Market analyst Ed Zitron went further, arguing that OpenAI and Anthropic simply don't have enough revenue to finance the massive data center construction costs they're projecting. Zitron estimates a truly sustainable AI industry might generate less than fifty billion dollars annually—a massive gap from current trillion-dollar valuations. If he's right, we should start seeing large-scale cancellations of data center projects as demand fails to materialize. The author suggests future AI development will increasingly move into classified government research, leaving the commercial sector operating at much smaller scale.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton