Imbue co-founder: alignment isn't a perfect-formal-system problem, it's risk-vs-investment
joshalbrecht · x · 2026-09-10
Josh Albrecht pushes back on the framing "we don't know how to align superintelligence." He argues no one seriously believes some formal system would "perfectly" align an agent.
If perfection is off the table, the question is empirical — and we already know a lot about doing alignment empirically for today's systems. Just as we can't make perfectly safe anything, he says alignment is fundamentally a ratio between risk and investment in empirical safety procedures, not a mystical binary.
More from AGI Musings
- Commentary: Anthropic's jobs report is unpredictable; ex-employee doom talk may be PR — ziv_ravid · 2026-09-10
- Commentary: ex-OpenAI/Anthropic staffer's AI doom exit and Anthropic's jobs report spin — MatthewChang · 2026-09-10
- AI Fear-Mongering Has Gone Mainstream: Normies Now Think LLMs Are Black Magic — bindureddy · 2026-09-10
- Your competition isn't AI but people lazily prompting it, says Paras Chopra — paraschopra · 2026-09-10
- Anthropic CEO Dario Amodei puts P(doom) at 10-25%, some say his remarks run higher — menhguin · 2026-09-10
- AI Doom Imagery Is Weirdly All Fauci, and One X Thread Lays Out 1000 Kill Paths — PaulYacoubian · 2026-09-10