Episode summary
Can Open Models Keep Up With OpenAI and Anthropic? | Nathan Lambert
The a16z Show
· · 59 min · Hosts: Lisha Li, Erik Torenberg · Guests: Nathan Lambert
Summary
Nathan Lambert argues that model progress is mostly a contest over compute, data, and talent, with OpenAI and Anthropic ahead because they have more of all three and better org structures to turn those resources into frontier models. He thinks Chinese labs are technically strong and highly motivated, but remain more compute-starved, so their “fast follower” strategy is to ship open models and close the gap through distillation, post-training, and commercialization.
He is skeptical that distillation alone explains the speed of open-model progress. In his view, distillation helps, but the bigger story is that frontier labs are buying high-quality data, building environments, and shaping post-training around RL-heavy workflows that are hard to observe from outside. He argues the most important gains are increasingly coming from data quality, evaluation design, and the ability to create worlds where models can act and be scored.
Lambert also lays out Trillium’s mission: release open-source environments, trajectories, and documentation that make AI research more legible, especially around recursive self-improvement. He expects open research to be slower than closed labs on the frontier, but still valuable because it can lower barriers, improve public understanding, and give startups and researchers a clearer base to build on.
Key takeaways
- Lambert says the frontier is driven by compute, data, and talent, and that OpenAI and Anthropic have the strongest combination of those resources.
- He argues Chinese model builders are technically excellent but materially constrained by compute, not talent.
- Lambert thinks distillation matters, but only as one part of a broader post-training stack that now depends heavily on RL environments and data quality.
- He expects open and closed model labs to operate on different time scales, with closed labs first to monetize high-value use cases like coding.
- Trillium will focus on releasing open environments and documentation so researchers can study RL, automation, and recursive self-improvement more directly.
Notable quotes
“When I think about model progress, it comes down to different forms of capital and that is compute, data, and talent”
“The Chinese are so good at building models. To say that China is anything but equal on talent, you're going to have a hard time”
“What they don't have is compute”
“People distill my products every single day. They take it, they strip it down to its raw parts. That's called competition”
“I think open and closed models are on different time scales of exponentials”
How the episode opens
When I think about model progress, it comes down to different forms of capital and that is compute, data, and talent. — How do you see that playing out in terms of like who's going to win in the next couple of years?
The Chinese are so good at building models. To say that China is anything but equal on talent, you're going to have a hard time. What they don't have is compute. — Most people regard distillation as sort of like the main lever for them to do this.
People distill my products every single day. They take it, they strip it down to its raw parts. That's called competition. Previous post training was much more reliant on the supervised fine-tuning and now this has morphed a bit into environments where it's RL heavy but it's just like hard to map this type of post training which is you get the completions and the score goes up to what is going on in reasoning today.
The discourse now is just so much fear on like how much the frontier is just hidden and also breaking away. The AI capability plot doesn't necessarily worry me, but the AI capability plot in the pretty much non-existing preparation of cross public cyber security and all these things like so Nathan really happy to have you here. Uh thank you for joining in person.
Super timely given you're launching Trillium which I'd love to talk about. um but more sort of on on a macro level um I think we chatted a lot about open source and that's you know a lot of what you care about and uh what I sort of um or what we've been hearing a lot in the public discourse is like hey we have to pace the frontier you know the frontier this the frontier that and I think an arc of the stuff that I'm really trying to ask e experts about is like what is in the frontier what is it even like what are we pacing what's there and then more importantly kind of threading it with open source and open weight models like can they keep up with the frontier?
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- open vs closed models
- compute constraints
- distillation debate
- data and rl environments
- recursive self-improvement
Companies and tickers mentioned
- OpenAI
- Anthropic
- Trillium
Watch the episode
Watch “Can Open Models Keep Up With OpenAI and Anthropic? | Nathan Lambert” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from The a16z Show
-
How AI Could Bring Halo and Other Classic Games Back to Life -
Mike Volpi on Why AI Infrastructure Will Outlast the Hype Cycle -
Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse -
Inside Salesforce’s Acquisition Strategy for the AI Era -
How AI Is Turning Figma’s Finance Team Into Builders -
The New AI Agents That Shop, Plan, and Gossip
The a16z Show, this episode and its artwork belong to Andreessen Horowitz. Octagon is not affiliated with the show; this page is an independent summary.
