Machine civs lack mortality's regularization — 'like training MoE without load balancing'
kellerjordan0 · x · 2026-09-09
keller Jordan explains his post series aimed to 'spook Landians': the standard argument (misaligned AI values) fails on Land-sympathetic camp who feel awe at AI transcending humanity. His science-fictional case: human civilizations are well-regularized by mortality and inability to copy ourselves; machine civs turn these controls off — so no one knows if a global machine civ is stable, 'like training MoE without load balancing.' An attempt at an AI-hesitancy argument compelling to both EA and Landian types.
More from AGI Musings
- Stanford-Harvard ARISE releases inaugural State of Clinical AI Report 2026 — jonc101x · 2026-09-09
- Anthropic researcher: without AI, US GDP growth would be ~1% — QuintinPope5 · 2026-09-09
- Ben Todd: if alignment experts say it could kill everyone, react with "oh shit", not PR critiques — sjgadler · 2026-09-09
- Is 'Pointing Enough ChatGPT Agents at Navier-Stokes' an Extreme Idea? — teortaxesTex · 2026-09-09
- "Change Could Come Any Day": Poster Distrusts OpenAI and Anthropic on Alignment — justalexoki · 2026-09-09
- tszzl: solving mechinterp within a year could halve his p(doom); utopia is seriously possible — tszzl · 2026-09-09