Frontier model training is a monopoly business: the bull case for Anthropic's 'One Ring' thesis
abhiadesai · x · 2026-11-03
Noahpinion lays out the bull case for Anthropic: intelligence is such a general technology that producing it lets you control all other technologies — the "One Ring to Rule Them All" thesis. abhiadesai agrees and extends: frontier model training is inherently a monopoly business, and "red queen race" criticisms underestimate capabilities growth — a 6-month training lead could enable hyper-adversarial behavior from the leader to eliminate competitors (citing OpenAI's "melt GPU" comment).
He offers a thought experiment: is it easier for the rocket company to compete with the superintelligence company, or vice versa? Counterintuitively, the latter is far easier.
More from AGI Musings
- Yoav Goldberg: the class of things we assume AI can now do is kind of insane — teortaxesTex · 2026-11-03
- "Reddit is now 80% bots": dev laments AI-driven collapse of information ecosystems — vboykis · 2026-11-03
- Zvi rebuts Scott Alexander: case against SURA strategy is 'deeply confused' — TheZvi · 2026-11-03
- 'Everyone is trained on copyrighted data' — another voice in the AI copyright debate — Merzmensch · 2026-11-03
- Vicki Boykis: my 20-year mental model of building products on the internet collapsed — vboykis · 2026-11-03
- Markets massively underprice AI if RSI is coming, argues pacing debate chart — ohlennart · 2026-11-03