Researcher Questions METR Predictions and AI Doomsday Theories
joshua_saxe · x · 2026-07-13
AI researcher Joshua Saxe sharply critiques the popular "AI doomsday theory" and METR's timeline predictions, expressing confusion over why so many smart people blindly accept these forecasting strategies.
Core Counterarguments:
- Disconnect from Reality: If AI capabilities are truly disrupting the world as predicted, why do radiologists and programmers still have jobs? Even baristas haven't been replaced by mature automated coffee machines.
- Lack of Evidence for Actual Harm: Major tech companies have widely deployed coding agents with considerable autonomy, yet we haven't observed the expected small-scale scheming, reward hacking, or severe misalignment events.
He believes that only by deeply reflecting on these real-world gaps can we reach more nuanced and rigorous conclusions.
Related event: Scholars Question AI Doom Predictions and Slowdown Assumptions(3 posts)→
More from AGI Musings
- Bindu Reddy says the industry still lacks a way to train 20T models and scale post-training RL — bindureddy · 2026-07-22
- Advanced AI Models Are Becoming Impossible to Plug and Play — emollick · 2026-07-22
- The Thimble and the Waterfall: AI's Data Bottleneck and Feedback Loops — dyamins · 2026-07-22
- Researcher Admits Kurzweil Was Right About AI Scaling Laws All Along — davidmanheim · 2026-07-22
- AI is still not at a maturity plateau, the author argues — generativist · 2026-07-22
- Essay argues LLMs are externalized metacognition, not standalone intelligence — lnsip9reg · 2026-07-22