Scott Aaronson: the wild AI prophecies came true — 'The Age of Wonders and Terrors'
_beenkim · x · 2026-09-17
Quantum computing scholar Scott Aaronson's essay 'The Age of Wonders and Terrors' lays out his updated stance on AI risk. Twenty years ago, the academic consensus was that superintelligence takeover would be preceded by warning signs: agents escaping containment, conspiring, solving Millennium Problems — panic only then. His current view: those prophecies are coming true, from AlphaGo through LLMs and reasoning agents to this summer's crescendo, and denying it requires "a particular kind of idiot." He skewers the endless shell game where skeptics move the goalposts each time AI achieves the previously impossible, arguing this is why the risks deserve serious attention now.
More from AGI Musings
- Oxford's Sandra Wachter on likelihood of an AI slowdown, cooperation and accountability — SandraWachter5 · 2026-09-17
- In an interview, engineer argues AI like Astra will automate most dev roles — ashishllm · 2026-09-17
- François Fleuret: AI math is to math what numerical integration is to ODEs — francoisfleuret · 2026-09-17
- Luca Ambrosio: AI has no nefarious intents, just blind obedience — LucaAmb · 2026-09-17
- Engineer's interview ended midway after admitting he hadn't used AI — CtrlAltDwayne · 2026-09-17
- Beyond "stochastic parrots": Wigner's unreasonable effectiveness now applies to LLMs — anselm · 2026-09-17