#979 - Dwarkesh Patel - AI Safety, The China Problem, LLMs & Job Displacement
Modern Wisdom
Chris Williamson
4.7 • 4.6K Ratings
🗓️ 11 August 2025
⏱️ 166 minutes
🧾️ Download transcript
Summary
Transcript
Click on a timestamp to play from that location
| 0:00.0 | What do you think that we've realized about human learning and human intelligence from architecting |
| 0:05.8 | AI intelligence? There's this really interesting thing we've seen where these AI models are |
| 0:14.8 | making progress first in the domains that we think of as the archetype of the where humans have their |
| 0:23.5 | primacy, right? So if you look at Aristotle, what does he say? What makes humans unique? |
| 0:28.3 | Well, it's reasoning. Humans can reason and other animals can't. And these models, these AI models, |
| 0:33.6 | they're just not that useful if you've tried to use them for your work. They're useful in certain |
| 0:37.0 | domains, but broadly they're just not widely deployable. What is the one thing that they can do? They can reason. But they obviously, they can't carry a cup of water, right? Robotics isn't solved. They can't even, like, do a job. They can't even do a white-collar job. So there's this interesting thing called Moravex paradox. Hans Morawak came up with this idea in the 90s, where he noticed that the tasks which are |
| 1:00.8 | easiest for humans are taking computers the longest to solve. So we still haven't solved robotics yet. It's so easy for us to move around. |
| 1:08.8 | Whereas the tasks which are quite hard for humans, |
| 1:10.9 | like adding numbers, adding long numbers. Computers could do that in the 60s. And the logic there |
| 1:15.6 | is that evolution has only optimized us for, let's say, the last million years to be good at |
| 1:22.8 | reasoning, to be good at arithmetic, to be good at these kinds of high-level abstractions. |
| 1:28.0 | And it was just to spend four billion years teaching us how to move around the world, |
| 1:32.3 | how to pursue your goals in a long-term basis, so not just do this task over the next hour, |
| 1:38.1 | but spend the next month planning how to kill this gazelle. |
| 1:42.7 | That has been, I think, remarkably accurate predictor of the places we've seen |
| 1:47.5 | in the air progress. |
| 1:48.2 | They're like, they're automating coding. |
| 1:50.3 | Coding we thought of was this thing that 0.1% of the population could do really well. |
| 1:54.1 | That's the first thing that went below the water line. |
| 1:58.0 | And, yeah, just like basic, you know, manual work might genuinely be the last thing that goes away. |
| 2:03.0 | Right. Yeah, there's a difficulty in getting a robot to crack an egg. |
... |
Transcript will be available on the free plan in 5 days. Upgrade to see the full transcript now.
Disclaimer: The podcast and artwork embedded on this page are from Chris Williamson, and are the property of its owner and not affiliated with or endorsed by Tapesearch.
Generated transcripts are the property of Chris Williamson and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.
Copyright © Tapesearch 2025.