ludwigABAP digs up 20 old 'collapse' essays and cringes at his own AGI doom predictions

ludwigABAP · x · 2026-11-20

Well-known AI commentator ludwigABAP searched his own archive and found 20 long essays he once wrote about "collapse" and the "1D" scenario, mocking himself for taking the discourse that seriously. The quoted old thread argued that context window expansion carries deep compute-cost implications needing research, and that domains like compiler internals, hardware-aware optimization, and large-scale distributed/concurrency engineering aren't being solved by agentic LLMs. The post doubles as a retrospective on those predictions.

Original post →

More from AGI Musings

AGI Musings channel →