Miles Brundage: 'iterative deployment' is often expediency dressed up as safety
Miles_Brundage · x · 2026-10-04
Former OpenAI safety VP Miles Brundage clarified he wasn't assuming bad faith in an ongoing argument — just impatient 'with a few hundred thousand minutes until superintelligence.' His core point: companies and people inside them are incentivized to make a virtue of necessity, and the term 'iterative deployment' is often used that way — glorifying rapid shipping as if it were the pro-safety move when it isn't.
He grants that AI is a general-purpose technology and reality is a good teacher, which is why the framing has 'gotten away' with it. But he argues we almost always already know more we could do to make each launch safer.
More from AGI Musings
- Steering the "pain direction" makes models choose irreversible harm 94% of the time — repligate · 2026-10-04
- Straight lines on graphs: you can't even see where AI happened — tszzl · 2026-10-04
- Would an LLM trained only on pre-1900 data predict a world war? — dbasch · 2026-10-04
- Nathan Lambert decries AI ecosystem norm of 'you're evil' attacks on safety work — novasarc01 · 2026-10-04
- AI code generation hits inflection point as synthetic data opportunities outpace ability to exploit them — mrjonfinger · 2026-10-04
- Commentator warns ZIRP-fueled abundance may unwind before AI supergrowth arrives — ericwdolan · 2026-10-04