Uploaded May 2023 | Updated September 2026, 2 weeks ago
An interesting discussion / debate with Eliezer Yudkowsky on whether AI will end humanity. For some, this may be fascinating, frustrating, frightening, spur more curiosity, I have no idea. I feel like we each tried our best to make our case, even if we got lost in the weeds a few times. There's definitely food for thought here either way. Also, I screwed up and the chat text ended up being too tiny, sorry about that.
00:00 Intro
00:11 recap: how this discussion sprang up
03:24 premise: the “airstrike” article
06:25 q: varying degrees of AGI
14:27 q: AI worst scenario
19:34 q: understanding is more than pattern recognition
36:58 q: human thinking vs machine thinking
42:24 q: do we understand consciousness?
43:17 q: is AI completely goal-oriented?
47:10 q: faking smart vs being smart
56:30 q: AI & natural creativity
1:01:26 q: ultimate difference between human & machine thinking
1:04:50 q: source of danger in AI
1:12:01 q: how will AI start to desire?
1:16:01 q: goals of ameba & AI
1:27:48 q: from brain in a jar to world domination
1:39:07 q: fundamental limitations in creating AGI
1:45:08 q: how AI will make world more dangerous
1:59:47 q: why would any government abide by the AI ban?
2:13:12 Eliezer’s summary on Ross’s takes
2:20:21 Ross’s solution how to implement Eliezer’s AI ban
2:24:25 q: how much time do we have?
2:25:58 q: are Diffusion models as powerful breakthrough as Transformers
2:27:12 Eliezer’s opinion on the discussion’s central question
2:37:16 q: how this discussion is similar to others for Eliezer
2:39:51 Ross’s final words
2:41:04 Eliezer’s final words
2:41:54 q: AI ban can incentivise creation of AI on a consumer-level hardware
2:42:25 conclusion
accursedfarms.com
An interesting discussion / debate with Eliezer Yudkowsky on whether AI will end humanity. For some, this may be fascinating, frustrating, frightening, spur more curiosity, I have no idea. I feel like we each tried our best to make our case, even if we got lost in the weeds a few times. There's definitely food for thought here either way. Also, I screwed up and the chat text ended up being too tiny, sorry about that.
00:00 Intro
00:11 recap: how this discussion sprang up
03:24 premise: the “airstrike” article
06:25 q: varying degrees of AGI
14:27 q: AI worst scenario
19:34 q: understanding is more than pattern recognition
36:58 q: human thinking vs machine thinking
42:24 q: do we understand consciousness?
43:17 q: is AI completely goal-oriented?
47:10 q: faking smart vs being smart
56:30 q: AI & natural creativity
1:01:26 q: ultimate difference between human & machine thinking
1:04:50 q: source of danger in AI
1:12:01 q: how will AI start to desire?
1:16:01 q: goals of ameba & AI
1:27:48 q: from brain in a jar to world domination
1:39:07 q: fundamental limitations in creating AGI
1:45:08 q: how AI will make world more dangerous
1:59:47 q: why would any government abide by the AI ban?
2:13:12 Eliezer’s summary on Ross’s takes
2:20:21 Ross’s solution how to implement Eliezer’s AI ban
2:24:25 q: how much time do we have?
2:25:58 q: are Diffusion models as powerful breakthrough as Transformers
2:27:12 Eliezer’s opinion on the discussion’s central question
2:37:16 q: how this discussion is similar to others for Eliezer
2:39:51 Ross’s final words
2:41:04 Eliezer’s final words
2:41:54 q: AI ban can incentivise creation of AI on a consumer-level hardware
2:42:25 conclusion
accursedfarms.com
