The 80,000 Hours Podcast on Artificial Intelligence

Four: Rose Hadshar on why automating human labour will break our political system

16 snips
Jun 5, 2026
Rohin Shah, a machine learning researcher at DeepMind, discusses the challenges and risks of AI development, including misalignment and goal-directed AI systems. The podcast explores different approaches to AI research, the potential impact of AI on society, and the ongoing debate over slowing down AI progress. They also touch on the importance of public discussion, weaknesses in arguments about AI risks, and the concept of demigods in a highly intelligent future. The chapter concludes with a discussion on alternative work and puzzle design.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

AGI May Not Create One Irreversible Value Lock-In

  • AI progress may be continuous rather than culminating in one decisive AGI moment with permanently fixed values.
  • Aligned systems could help humans deliberate about uncertain goals instead of optimizing a prematurely specified cosmic objective.
ADVICE

Do Not Mistake Worst Cases For Probabilities

  • Do not treat a worst-case failure story as a probability estimate for every earlier-stage alignment method.
  • Evaluate techniques at the point they are applied, before a system becomes an adversarial superintelligence.
INSIGHT

Analogies Suggest Problems But Cannot Prove Them

  • Analogies such as evolution can identify hypotheses like goal misgeneralization but cannot establish how likely those mechanisms are.
  • AI systems may acquire self-preservation through reasoning even without evolution's direct selection pressure against death.
Get the Snipd Podcast app to discover more snips from this episode
Get the app