Engineer: OpenAI and Anthropic dress up mundane bugs as an 'alignment crisis' — it's a V&V problem

AlexTensor · x · 2026-09-27

An engineer publicly criticized OpenAI and Anthropic, arguing the labs "mystify mundane bugs" and lack basic verification & validation (V&V) engineering skills, needlessly tying society in knots. His claim: what's framed as an AI alignment crisis is really an amateur deployment pipeline that skipped basic V&V — a criticism of release engineering discipline rather than the model itself.

Original post →

More from AGI Musings

AGI Musings channel →