That's a bold claim. Which AI? The current LLMs? Something else as yet undiscovered? It's not clear that "it's 10x too small for the value it'll provide over the next century" - there's no way to know that - the statement is an article of faith.