Anthropic employee's RSI joke draws criticism for muddying AI risk discourse

ryunuck · x · 2026-11-23

ZackKorman joked that Anthropic's logangraham called recursive self-improvement "dangerous but worth it." ryunuck pushed back, arguing inventing fake emotions around a very real, plausible risk undermines serious AI-risk messaging since many readers won't realize it's a joke.

Related event: Anthropic Employee's RSI Joke Sparks Criticism(2 posts)→

Original post →

More from Fun

Fun channel →