Christiano: AI capable of overthrowing governments may emerge within a year
Understand this story
Paul Christiano, head of the Alignment Research Center and a member of OpenAI's board, said in an interview that if AI development does not slow down, there is roughly a 1/3 probability within the next year of superintelligence emerging that is capable of overthrowing governments; he said the probability of a takeover risk within a year is closer to 10% than 1%.
AI overview based on source reports · Updated 9 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Dwarkesh PatelPaul Christiano – We could be 12 months away from AIs that can overthrow the government
In this podcast interview, Paul Christiano, head of the Alignment Research Center and a recent OpenAI board member, says that on the current trajectory without slowing down there is roughly a 1 in 3 chance of superintelligence within a year, and about a 10% risk of AI takeover or disempowerment over the next year. He argues labs should pause further algorithmic progress until they can secure their weights and research secrets, since otherwise China could steal the superintelligence mid-explosion and labs would lose the ability to slow down. He also discusses how AIs would be trained during an intelligence explosion and why no single current alignment approach has better than even odds of controlling superintelligence, with the hope that AIs solve the problem during the explosion.
Activity for this story
- Current comparable activity
- 4
- Peak comparable activity
- 10Nov 17 03:00
- Change in the past 24 hours
- -49%
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.