Timothy Lee: AI takeover risk over decades isn't negligible, especially with robotics

binarybits · x · 2026-11-16

In a debate with Jess Riedel and Tom Dietterich, Timothy Lee (binarybits) argues that on a timescale of decades, AI takeover risk cannot be called negligible — especially once robotics enters the picture: "a lot of weird stuff could happen." Riedel counters that the disagreement boils down to whether returns to intelligence above adult-human level diminish so fast that takeover risk is negligible — plausible near-term, implausible long-term.

Related event: Researchers debate diminishing returns of superhuman AI and takeover risk(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →