Nathan Lambert

speaker

1,814 appearances 3 recordings 2 series first heard Feb 2025 last heard 1 Feb

Nathan Lambert’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 2 in all, peaking in Feb 2026 with 1.

Appearances

newest first · ▶ plays the moment

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

And it's just like that set of secure cloud environment is just so nice for just like

send it off and do this thing and then come back to me.

And these will probably help define some of the local open and closed niches.

But I think initially, because there was such a rush to get this tool use working, that the open models were on the back foot.

which is kind of inevitable.

I think there's so many resources in these frontier labs, but it will be fun when the open models solve this because it's going to necessitate a bit more flexible and potentially interesting model that might work with this recursive idea to be an orchestrator and a tool-used model.

So hopefully the necessity drives some interesting innovation there.

This relates a lot to this kind of SF zeitgeist of what is AGI, which is artificial general intelligence, and what is ASI, artificial super intelligence, and what are the language models that we have today capable of doing?

I think the language models can solve a lot of tasks, but a key milestone among the AI community is essentially when AI could replace any remote worker, taking in information and solving digital tasks and doing them.

And the limitation that's highlighted by people is that a language model will not learn from feedback the same way that an employee is.

So if you hire an editor, the editor will mess up, but you will tell them.

And if you hired a good editor, they don't do it again.

But language models don't have this

ability to modify themselves and learn very quickly.

So the idea is if we're going to actually get to something that is a true general adaptable intelligence that can go into any remote work scenario, it needs to be able to learn quickly from feedback and on-job learning.

I'm personally more bullish on language models by being able to just provide them with very good context.

You said, like you maybe offline said that

like you can write extensive documents to models where you say, I have all this information.

Here's all the blog posts I've ever written.

I like this type of writing.

Showing 761–780 of 1,814 · page 39 of 91