Challenging the AI takeover threat model: alignment is reasonably easy
teortaxesTex · x · 2026-11-16
teortaxesTex challenges the dominant "AI takeover" threat model: if takeover were the real concern, you'd expect a substantial alignment tax and an unfettered agent able to beat both humans and aligned AIs despite a massive resource gap — the old LessWrong doctrine. The current frontier mindset, he argues, is instead "lmao Mythos clears, look at our gigawatts of capacity." There are no Forbidden Magic shortcuts or greater daemons to summon while skipping the binding rite; alignment is reasonably easy, and people only care about power controlled by people — what he calls "ape power centralization."
More from AGI Musings
- Conference-goer memes 'we're banning RSI' into existence at Takeoff — austinc3301 · 2026-11-16
- Dwarkesh's 'Recursive self-improvement has started' hits 539K views in 2 days — shashib · 2026-11-16
- If AI replaces junior work, firms still need a path to build reviewers' judgment — TansuYegen · 2026-11-16
- Percy Liang to AI labs: share non-vacuous info outsiders can actually evaluate — lateinteraction · 2026-11-16
- New Interview: Stephen Wolfram on the Ruliad and Why We Can't Predict It — romanyam · 2026-11-16
- Dan Robinson: AI likely able to cause $1T damage before it can fully do your job — austinc3301 · 2026-11-16