High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / belief

Published · transcript-backed

Nathan Lambert: belief

1 Feb 2026 Lex Fridman Podcast #490 – State of AI in 2026: LLMs, Coding, Scaling Laws, China, Agents, GPUs, AGI

“I think eventually they’ll fool you, and it’ll be on platforms that give ways of verifying or building trust.”

— Nathan Lambert

Source trail

Everything needed to verify it.

Speaker
Nathan Lambert
Attribution
Verified speaker
Claim type
belief
Recorded
1 Feb 2026
Publisher
Lex Fridman Podcast

Transcript context

…Even like classic examples, I honestly think this is true, and I think we will get tired of it. We are already kind of tired of it. Same with art. I don’t think art will go away. I mean, you have physical paintings. There’s more value, not just monetary value, but just more appreciation for the actual painting than a photocopy of that painting. It could be a perfect digital reprint, but there is something when you go to a museum and you look at that art and you see that real thing and you just think about, “Okay, a human.” It’s like a craft. You have an appreciation for that. And I think the same is true for writing, for talking, for any type of experience, where it will be… I do unfortunately think it will be like a dichotomy, like a fork where some things will be automated. There are not as many paintings as there used to be 200 years ago. There are more photographs, more photocopies. But at the same time, it won’t go away. There will be value in that. I think that the difference will just be what’s the proportion of that. But personally, I have a hard time reading things where I see it’s obviously AI-generated. I’m sorry, there might be really good information there, but I have a certain feeling, like, it’s not for me. I think eventually they’ll fool you, and it’ll be on platforms that give ways of verifying or building trust. So you will trust that Lex is not AI-generated, having been here. So then you have trust in this- -channel. But it’s harder for new people- -that don’t have that trust. Well, that will get interesting because I think fundamentally it’s a solvable problem by having trust in certain outlets that they won’t do it, but it’s all going to be kind of trust-based. There will be some systems to authorize, “Okay, this is real. This is not real.” There will be some telltale signs where you can obviously tell this is AI-generated and this is not. But some will be so good that it’s hard to tell, and then you have to trust. And that will get interesting and a bit problematic.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.

Search evidence