All Segments
Eliezer Yudkowsky: "AI alignment researchers aren't productive"
4 segments available
From the debate between Eliezer Yudkowsky & George Hotz. Watch the full vod here: https://youtube.com/live/6yQEA18C-XI Twitter: https://twitter.com/dwarkesh_sp Apple Podcasts: https://apple.co/3oBack9 Spotify: https://spoti.fi/3S5g2YK
Segments Timeline
The Urgency of AI Alignment
Eliezer Yudkowsky discusses the critical nature of AI alignment and questions the significance of the speed at which AI development occurs. He suggests that if the threat is imminent, the pace of progress may not matter, but acknowledges that a slower pace could provide an opportunity to address alignment issues.
"If we're just as doomed if it goes slowly, then what does it matter if it goes slowly? Wait, wait, wait, wait, wait, wait. What? Does it matter if it goes slowly or quickly? If it goes slowly, w..."
The Urgency of AI Alignment
In this segment, Eliezer Yudkowsky discusses the critical nature of AI alignment and questions the significance of the speed at which AI development occurs. He suggests that if the threat is imminent, the pace of progress may not matter, but if it slows down, there might be an opportunity to address the alignment problem.
"If we're just as doomed if it goes slowly, then what does it matter if it goes slowly? Wait, wait, wait, wait, wait, wait. What? Does it matter if it goes slowly or quickly? If it goes slowly, w..."
Productivity in AI Research
Yudkowsky expresses skepticism about the productivity of current AI alignment researchers. He argues that the effectiveness of their thinking is more crucial than the amount of time they have to solve the problem, indicating a lack of confidence in their approaches to AI alignment.
"Which problem? AI alignment. Oh, that one? Haha. Um, that, yeah. It's not gonna go that slowly. I'm not so sure. Okay. I, I've, I've looked at these people trying to solve this thing and I'm n..."
Productivity vs. Time in AI Research
Yudkowsky emphasizes that the effectiveness of AI alignment researchers is not determined by the amount of time they have but rather by the productivity of their thinking. He expresses skepticism about the current efforts to solve the alignment problem, suggesting that time alone will not lead to solutions without productive thought.
"Which problem? AI alignment. Oh, that one? Haha. Um, that, yeah. It's not gonna go that slowly. I'm not so sure. Okay. I, I've, I've looked at these people trying to solve this thing and I'm n..."