AI Doom Theories Called Ridiculous, Human Writing Framed as Costly Signal of Intelligence
zetalyrae · x · 2026-09-25
Citing a podcast with a Redwood Research AI safety expert, Isaac King argues current AI takeover theories rest on dozens of contingent assumptions with made-up probabilities—change one variable and the doom web unravels. Extraordinary extinction claims shouldn't drive policy without extraordinary evidence.
He adds that human writing is a costly signal of intelligence: AI writing isn't morally wrong and is increasingly pleasant, but people with things worth saying correlate with those who want to say it themselves.
Related event: Redwood researchers' AI takeover theories draw criticism(2 posts)→
More from AGI Musings
- Investor Stewart Alsop: AI agents are shattering an already fragile social trust — StewartalsopIII · 2026-09-25
- BlackRock says AI agents could soon use stablecoins to buy compute and data autonomously — Polymarket · 2026-09-25
- Grady Booch: "artificial intelligence" predates the first print use of "software" — Grady_Booch · 2026-09-25
- The Body Remains: why embodied knowledge matters in the agentic age — DrKavner · 2026-09-25
- Princeton professor: stale Wikipedia pages spread misinformation about not-quite-famous experts — random_walker · 2026-09-25
- SEO is giving way to Agent Optimization, and scientific papers may follow — CSProfKGD · 2026-09-25