Dan Roberts

Scientist @OpenAI. Prev. co-founder @diffeo, acquired by @salesforce // co-authored The Principles of Deep Learning Theory // studied gravity.

[Dan Roberts @danintheory]

TL; DW I say things (on reinforcement learning, the process of scientific discovery, and how physicists approach work on AI).

[Matt Turck @mattturck]

Why AI Can Now Make Discoveries - my conversation with @danintheory, Lead of the Foundations of Reinforcement Learning team at @OpenAI

00:00 Intro: AI's wild week in mathematics

01:21 What OpenAI's Foundations of RL team does

03:08 Dan's journey: from black holes and quantum gravity to frontier AI

07:04 Are AI systems becoming useful for real science

08:21 The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic

08:52 Why the OpenAI result was an act of exploration

10:25 OpenAI vs. DeepMind: informal reasoning vs. formal proof

12:13 RL 101: learning by doing, not just watching

15:10 Why reinforcement learning works

15:58 How RL breaks: sparse feedback and long-horizon tasks

17:03 RLHF: how human feedback shaped early language models

18:48 Move 37, self-play, and the search for novel strategies

22:16 Explore vs. exploit in scientific discovery

24:49 Why RL may now be "the cake," not the cherry on top

25:46 Why RL started working with large language models

27:29 Is RL "sucking supervision through a straw"?

28:47 Why language may be the grounding layer for intelligence

31:46 A contrarian take on the Bitter Lesson

32:41 What test-time compute actually is

34:50 How RL gives models the ability to think

35:40 Verifiable rewards, math, coding, and the messy real world

38:00 What physics can teach us about AI

42:08 Is there a thermodynamics of AI?

43:08 From Erdős problems to Einstein-level AI

45:16 Is AI already doing original science?

45:51 How far are we from AI automating AI research

47:41 Why Dan is excited about the future of science

[Dan Roberts @danintheory]

@carterjwilcox Crushing it :)

[Dan Roberts @danintheory]

I’m excited for researchers to come up with new and inventive ways of consuming these inference wins…

[Etched @Etched]

We're coming out of stealth.

We've built our first racks after a successful A0 tapeout, $1B+ in customer contracts, and $800m raised.

Early customer tests show us achieving SOTA throughput, latency, and power efficiency on inference workloads.

Our first racks ship this summer.

[Dan Roberts @danintheory]

@Etched Congratulations!!!

[Ryan Brewer @ryanbrewer]

Guy who hasn’t tried computer use on codex yet

[Noam Brown @polynoamial]

GPT-5.6 is incredibly strong and fast for coding. I hope we can make it available to everyone soon.

[OpenAI @OpenAI]

Sol is our new flagship and a step function better than GPT-5.5.

Terra delivers performance competitive to GPT-5.5 at 2x lower cost.

Luna is our most cost-efficient model, delivering strong capability at our lowest cost.

Together, the GPT-5.6 family gives people and developers more choice in how they balance intelligence, speed, and cost.

[Boris Hanin @BorisHanin]

I’d say I agree but a I’m a little biased 😇

[Dan Roberts @danintheory]

@BorisHanin lol what is that thread even?

[Dan Roberts @danintheory]

Great conversation and a fun way to learn about an important open AI problem!

[Sequoia Capital @sequoia]

Today's AI models train once. We don't work that way. We learn continuously, forget what doesn't matter, and retain what does.

That gap is what @dan_biderman and @realJessyLin are closing at @EngramLab. AI that never stops learning, with memory that lives inside the model instead of bolted on as an afterthought.

In our latest Training Data episode we get into why memory is the next frontier: why the brain forgets on purpose, why RAG is a band-aid, and what becomes possible when a model is always training.

00:00 Introduction
00:59 Always Training Explained
01:51 Beyond Context Windows
03:29 Ngram Product Overview
04:34 Adapters And Training Signals
05:32 Internalize Vs Externalize
06:49 Compute And Token Savings
08:19 Teams First Then Individuals
08:51 Memorization Vs Understanding
12:47 Dreams And Offline Digestion
14:08 Training Beats Curation
15:19 Why Everyone Needs A Model
21:44 Bitter Lesson And Architecture
24:44 RAG Killer And KV Cache
31:38 Future Of Memory And Models

[Dan Roberts @danintheory]

@bneyshabur @mirendil @a16z @kleinerperkins Congratulations!!!!

[Dan Roberts @danintheory]

Science and technology are great, and so is @bneyshabur and his team at @mirendil! Congratulations!

[Behnam Neyshabur @bneyshabur]

Today, I’m excited to formally announce @mirendil with my amazing co-founders Harsh Mehta, Shayan Salehian, and Tara Rezaei!

We’re fortunate to work with @a16z and @kleinerperkins, who led our seed round of $200M, followed by a major investment from NVIDIA, among others.

Mirendil exists to accelerate science and technology, and through them, to help solve humanity's most pressing problems.

Self-accelerating AI R&D is the most direct path to delivering on AI's broader promise, which is why we believe the most important application of AI is AI itself. Get this loop right, and it compounds. It fundamentally changes the rate of progress itself across all domains.

We believe this capability should be democratized. It should be used to power all scientific efforts trying to innovate at the frontier. There are far more important problems—and broader ones—than any single lab can take on, so more groups should be able to pursue them.

This pulls concentration of power away from a few labs: businesses and science labs can own their AI and infrastructure, keep their margins, and control their own destiny instead of ceding it all to a single AI lab.

We’re a small team with a singular focus. Our founding team consists of 20 researchers and engineers from frontier institutions including Anthropic, xAI, Google DeepMind, and OpenAI, united by a passion for science and a drive to build the technologies that move it faster. If you want to build the system that builds systems, join us!

@HarshMeh1a, @shayan_, @tararezaeikh

[Dan Roberts @danintheory]

Engram is making a personal AI that really learns and internalizes everything you care about. It’s a different kind of lab with an amazing team put together by @dan_biderman @EyubogluSabri @realJessyLin and @jxmnop. Very excited for what they are building!

[Engram @EngramLab]

https://t.co/CGIef5lIBI

[Dan Roberts @danintheory]

@EngramLab Congratulations!!!!

[Dan Roberts @danintheory]

@jmbollenbacher Sorry, but your post is definitely nonsense.

[Dan Roberts @danintheory]

@deanwball @SebastienBubeck @OpenAI Congratulations and welcome!!

[Dean W. Ball @deanwball]

I am pleased and honored to announce that, on July 6, I'll be joining @OpenAI as leader of a new team called Strategic Futures. Our mandate will be to help the company's leadership shape frontier AI policy. There is a ton of work to do, and I'm excited to get started.

[Sam Altman @sama]

We offer no explanation as to why Noams are so good at AI; we attribute their success, as all else, to divine benevolence.

[Dan Roberts @danintheory]

Not to mention the additional word play opportunities for no one…

[Dan Roberts @danintheory]

It’s very fun that “Noam at OpenAI” no longer disambiguates :)

[Noam Shazeer @NoamShazeer]

I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there.

It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you.

[Dan Roberts @danintheory]

@NoamShazeer Congratulations / welcome!