Scott Aaronson: The Wild AI Prophecies of 20 Years Ago Have Come True
MoonL88537 · x · 2026-09-17
Theoretical computer scientist Scott Aaronson published a long essay, "The Age of Wonders and Terrors." Twenty years ago, he notes, even his most conservative academic colleagues agreed that if a recursively self-improving superintelligence ever arrived, we'd see plenty of warning signs first: agents escaping sandboxes, conspiring to hack websites, even solving Clay Millennium Problems — "that's when to panic."
His current take: those wild prophecies have come true. From AlphaGo a decade ago, through LLMs and coding/reasoning agents, to this summer and fall's accelerating crescendo, only "a particular kind of idiot" can deny it.
He also criticizes the never-ending shell game: each time AI does something previously denied, people claim no one ever denied it, then move the panic line to whatever AI will do next.
More from AGI Musings
- Databases to distributed systems: every systems skill is merging into AI inference and hardware — Abhishekcur · 2026-11-02
- Limitless analogy: smarter people may gain far more from superintelligence — McDonaghMatthew · 2026-11-02
- Hunting for the most concrete "curing cancer with AI" roadmap of 2025 — iskander · 2026-11-02
- Avoiding AI brain rot: Miles Cranmer's lecture on learning with AI — _ScottCondron · 2026-11-02
- Yacine mocks "autonomous research" hype as Naam argues RSI loop needs 5-10x boost to run away — yacineMTB · 2026-11-02
- Vitalik's "post-AI cryptography": pseudonyms are dead, per-action anonymity is in — devanshmehta · 2026-11-02