Guest
Tom Davidson
Senior research fellow, Forethought
Co-authored research on how a small group could use AI to seize power, elevating AI-enabled coups as a risk.
Real Alpha has cut 6 moments of Tom Davidson (Senior research fellow, Forethought) talking, from an interview on 80,000 Hours Podcast published Sep 29, 2026.
What Tom Davidson said
80,000 Hours Podcast · AI Overlords vs Power-hungry Humans: Which Should Scare You More?
6 moments · 30 min
- 51:04
A pause needs safeguards against government power grabs Pro pick
“let's like properly red team the plan for pausing it, assuming a worst case scenario about the government actors that they kind of inherit that new regulatory environment.”
at 55:44 - 19:35
AI power grabs may provide clearer warning signs
“From the perspective of an AI collective across all time and space that just really wants to seize power, the Hugging Face Incident was a complete disaster.”
at 20:48 - 1:04:40
Two or three AI projects could balance safety and power
“you get a lot of benefit of power distribution from just a kind of a fairly small number of projects.”
at 1:08:07 - 42:40
A reward-seeking AI may not entrench its takeover
“if you're a reward seeking AI, it's a bit less clear that when you seize power and you seize control of the data center”
at 45:38 - 36:14
A human ruler may ignore advice to reconsider their values
“my worry is that there's no actually compelling argument that you could give to someone like this that they need to reflect”
at 41:30 - 1:22:32
Evidence about AI company control could shift their views
“And, you know, I think it's quite plausible that that shifts me, you know, to thinking that, yeah, the misalignment risk is notably bigger overall.”
at 1:24:06
- 51:04
Is this you? Add the badge to your site.
