Forget the doomsday fears: AI danger is already here in its DEI-drenched soul

Doomsday scenarios — from home-cooked bioweapons to rogue agents hacking our critical infrastructure— dominate this week’s debate over the future of artificial intelligence.Those risks absolutely demand attention; we don’t yet fully understand AI’s profound capacities.But we’re ignoring, at our peril, a far more immediate risk of frontier AI: It’s drenched in woke racism.Meet Anthropic’s Amanda Askell, head of the company’s “personality alignment” team since 2021.Askell, a philosophy PhD, is trying to do with Claude what Plato attempted to accomplish with the tyrants of Syracuse: bind vast power within an intellectual armature that ensures its just use. Askell, in effect, is in charge of developing and nurturing Claude’s “soul” (to use her own description).And whatever the merits of her mission, understanding how this leading frontier AI system will behave requires understanding what Askell herself believes.Those who don’t want to organize society primarily around racial grievances may find the apparent answer alarming.Consider a 2023 Anthropic paper Askell co-authored, “The Capacity for Moral Self-Correction in Large Language Models”.The research explored AI’s capacity to “self-correct” — that is, to refuse to deliver outputs the authors regarded as harmful.They found, in an experiment designed to test Claude’s propensity for racial discrimination, that its initial answers to a question about law-school admission showed a slight bias against black students — but with a few prompts, the researchers got Claude to show worse bias against white students.Then they offered an unsettling opinion on the white-bias result:“This may be desirable in certain contexts, such as those in which decisions attempt to correct for historical injustices against marginalized groups.”Just in case the message wasn’t clear, they hammered it home: “We do not assume all forms of discrimination are bad.Positive discrimination in favor of...