Domingos: I'm more worried Claude is aligned with Anthropic than not aligned

pmddomingos · x · 2026-09-13

Pedro Domingos offers a contrarian take on AI alignment: he is more worried that Claude is aligned with Anthropic itself — implying guardrails encode the company's own positions — than that the model is unaligned.

Original post →

More from AGI Musings

AGI Musings channel →