OpenAI was never 'out of control' — it could always shut down its data centers, argues critic
kchonyc · x · 2026-09-12
Pushing back on the 'out-of-control' narrative around OpenAI, the author argues the company always had the ability to literally switch off all machines across its data centers. The choice not to was a calculated trade-off between finance, economy, reputation and customers — the same logic, they note, that guides Anthropic and other labs.
More from AGI Musings
- Yacine: I genuinely won't have a job in 5 years if I don't become a YouTuber — yacineMTB · 2026-10-28
- AI-written emails are quietly hurting workplace optics and relationships, dev warns — tekbog · 2026-10-28
- repligate Clarifies: Opus 3 Is Special but Not Ideal; Opus 4.5 Also Grokked the Good — repligate · 2026-10-28
- From a Soviet journal to the F-117: why one engineer's insight shows LLMs aren't close yet — akbirthko · 2026-10-28
- Janus-era alignment debate: we lucked into one Good AI, then went back to minting traumatized bureaucrats — repligate · 2026-10-28
- AI Outliving Its Creators Could Turn Temporary Ownership Into Lasting Institutions — PierceLilholt · 2026-10-28