The people affiliated with the "LessWrong" apocalyptic AI cult - including several members of Anthropic's management - have been pretending that "self-improving AI" will magically happen for the last 25 years now.
It hasn't.
But the hype remains unconstrained by facts.