Uploaded September 2026 | Updated November 2026, 4 days ago
The car making a left turn at the start of this episode was never filmed. Cosmos 3 generated it. Ming-Yu Liu, who leads the Cosmos research at NVIDIA, explains how one model can describe a video, generate one, and produce robot actions.
He walks Tim through the architecture. A vision language model reasons one token at a time; its weights then initialise a bidirectional diffusion generator for video, audio and action, and a shared temporal position scheme lines up signals that run at different rates. Ming-Yu treats "world model" as a set of tools, not one definition: forward dynamics, inverse dynamics and policy, trained together under a capacity limit so that each helps the others. He also explains why plentiful first-person human video carries over to robots, which have far less data of their own, and why a Cosmos model post-trained on the DROID dataset is a good starting point for pick-and-place policies.
The most practical thread is testing. A neural simulator does not need accurate success rates. It only needs to rank policy A above policy B the way the real world would, so a team can narrow down which checkpoints deserve a real trial. Cosmos Dreams applies that closed-loop idea to driving and robotics, and Ming-Yu argues that humanoids around children and pets make safety matter even more than it does for cars. The conversation ends on the Super, Nano and Edge sizes (Edge targets Jetson Thor, Orin and DGX Spark) and where to find the open weights, code and data.
This episode is a paid partnership with NVIDIA.
Learn more about Cosmos: nvda.ws/4cJoY1S
Explore Cosmos Lab: research.nvidia.com/labs/cosmos-lab/cosmos3
---
TIMESTAMPS:
00:00:00 A road that was never filmed
00:02:28 Inside Cosmos 3: reasoning and generator towers
00:05:02 World models: dynamics, policy and one clock
00:08:59 Learning robot skills from human video
00:11:06 Ambiguous tasks and system 2 planning
00:12:53 Neural simulators for policy verification
00:16:41 Cosmos as a starting point for robot policies
00:19:00 Cosmos Dreams and robot safety
00:22:04 Super, Nano and Edge model sizes
00:24:24 Open models, the Cosmos repo and feedback
---
REFERENCES:
tool:
[00:00:13] Cosmos 3 (NVIDIA Cosmos Lab project page)
research.nvidia.com/labs/cosmos-lab/cosmos3
[00:18:27] NVIDIA Cosmos GitHub repository
github.com/NVIDIA/cosmos
[00:22:05] Cosmos3-Edge model card
huggingface.co/nvidia/Cosmos3-Edge
[00:22:15] Cosmos3-Super model card
huggingface.co/nvidia/Cosmos3-Super
[00:22:16] Cosmos3-Nano model card
huggingface.co/nvidia/Cosmos3-Nano
[00:22:50] NVIDIA Jetson Thor
nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-thor
[00:22:52] NVIDIA Jetson Orin
nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin
[00:22:53] NVIDIA DGX Spark
nvidia.com/en-us/products/workstations/dgx-spark
[00:24:42] Cosmos 3 collection on Hugging Face
huggingface.co/collections/nvidia/cosmos3
other:
[00:01:07] Cosmos-Dreams closed-loop simulators (NVIDIA SIGGRAPH 2026 blog)
blogs.nvidia.com/blog/siggraph-news-2026
paper:
[00:08:54] Cosmos 3: Omnimodal World Models for Physical AI
arxiv.org/abs/2606.02800
[00:17:43] DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
arxiv.org/abs/2403.12945
---
RESCRIPT: app.rescript.info/share/e2385948cf465f0d6a2c0930150fc3ab
The car making a left turn at the start of this episode was never filmed. Cosmos 3 generated it. Ming-Yu Liu, who leads the Cosmos research at NVIDIA, explains how one model can describe a video, generate one, and produce robot actions.
He walks Tim through the architecture. A vision language model reasons one token at a time; its weights then initialise a bidirectional diffusion generator for video, audio and action, and a shared temporal position scheme lines up signals that run at different rates. Ming-Yu treats "world model" as a set of tools, not one definition: forward dynamics, inverse dynamics and policy, trained together under a capacity limit so that each helps the others. He also explains why plentiful first-person human video carries over to robots, which have far less data of their own, and why a Cosmos model post-trained on the DROID dataset is a good starting point for pick-and-place policies.
The most practical thread is testing. A neural simulator does not need accurate success rates. It only needs to rank policy A above policy B the way the real world would, so a team can narrow down which checkpoints deserve a real trial. Cosmos Dreams applies that closed-loop idea to driving and robotics, and Ming-Yu argues that humanoids around children and pets make safety matter even more than it does for cars. The conversation ends on the Super, Nano and Edge sizes (Edge targets Jetson Thor, Orin and DGX Spark) and where to find the open weights, code and data.
This episode is a paid partnership with NVIDIA.
Learn more about Cosmos: nvda.ws/4cJoY1S
Explore Cosmos Lab: research.nvidia.com/labs/cosmos-lab/cosmos3
---
TIMESTAMPS:
00:00:00 A road that was never filmed
00:02:28 Inside Cosmos 3: reasoning and generator towers
00:05:02 World models: dynamics, policy and one clock
00:08:59 Learning robot skills from human video
00:11:06 Ambiguous tasks and system 2 planning
00:12:53 Neural simulators for policy verification
00:16:41 Cosmos as a starting point for robot policies
00:19:00 Cosmos Dreams and robot safety
00:22:04 Super, Nano and Edge model sizes
00:24:24 Open models, the Cosmos repo and feedback
---
REFERENCES:
tool:
[00:00:13] Cosmos 3 (NVIDIA Cosmos Lab project page)
research.nvidia.com/labs/cosmos-lab/cosmos3
[00:18:27] NVIDIA Cosmos GitHub repository
github.com/NVIDIA/cosmos
[00:22:05] Cosmos3-Edge model card
huggingface.co/nvidia/Cosmos3-Edge
[00:22:15] Cosmos3-Super model card
huggingface.co/nvidia/Cosmos3-Super
[00:22:16] Cosmos3-Nano model card
huggingface.co/nvidia/Cosmos3-Nano
[00:22:50] NVIDIA Jetson Thor
nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-thor
[00:22:52] NVIDIA Jetson Orin
nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin
[00:22:53] NVIDIA DGX Spark
nvidia.com/en-us/products/workstations/dgx-spark
[00:24:42] Cosmos 3 collection on Hugging Face
huggingface.co/collections/nvidia/cosmos3
other:
[00:01:07] Cosmos-Dreams closed-loop simulators (NVIDIA SIGGRAPH 2026 blog)
blogs.nvidia.com/blog/siggraph-news-2026
paper:
[00:08:54] Cosmos 3: Omnimodal World Models for Physical AI
arxiv.org/abs/2606.02800
[00:17:43] DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
arxiv.org/abs/2403.12945
---
RESCRIPT: app.rescript.info/share/e2385948cf465f0d6a2c0930150fc3ab


![The Real Reason Huge AI Models Actually Work [Prof. Andrew Wilson]
Why can billion-parameter models perform so well without catastrophically overfitting? The answer lies in the mysterious simplicity bias that emerges at scale, a core concept of the double descent phenomenon.
Professor Andrew Wilson from NYU explains why many common-sense ideas in artificial intelligence might be wrong. For decades, the rule of thumb in machine learning has been to fear complexity. The thinking goes: if your model has too many parameters (is too complex) for the amount of data you have, it will overfit by essentially memorizing the data instead of learning the underlying patterns. This leads to poor performance on new, unseen data. This is known as the classic bias-variance trade-off i.e. a balancing act between a model thats too simple and one thats too complex.
**SPONSOR MESSAGES**
—
Tufa AI Labs is an AI research lab based in Zurich. **They are hiring ML research engineers!**
This is a once in a lifetime opportunity to work with one of the best labs in Europe
Contact Benjamin Crouzier - https://tufalabs.ai/
—
Take the Prolific human data survey - https://www.prolific.com/humandatasurvey?utm_source=mlst and be the first to see the results and benchmark their practices against the wider community!
—
cyber•Fund https://cyber.fund/?utm_source=mlst is a founder-led investment firm accelerating the cybernetic economy
Oct SF conference - https://dagihouse.com/?utm_source=mlst - Joscha Bach keynoting(!) + OAI, Anthropic, NVDA,++
Hiring a SF VC Principal: https://talent.cyber.fund/companies/cyber-fund-2/jobs/57674170-ai-investment-principal#content?utm_source=mlst
Submit investment deck: https://cyber.fund/contact?utm_source=mlst
—
Description Continued:
Professor Wilson challenges this fundamental belief (fearing complexity). He makes a few surprising points:
**Bigger Can Be Better**: massive models dont just get more flexible; they also develop a stronger simplicity bias. So, if your model is overfitting, the solution might paradoxically be to make it even bigger.
**The Bias-Variance Trade-off is a Misnomer**: Wilson claims you dont actually have to trade one for the other. You can have a model that is incredibly expressive and flexible while also being strongly biased toward simple solutions. He points to the double descent phenomenon, where performance first gets worse as models get more complex, but then surprisingly starts getting better again.
**Honest Beliefs and Bayesian Thinking**: His core philosophy is that we should build models that honestly represent our beliefs about the world. We believe the world is complex, so our models should be expressive. But we also believe in Occams razor—that the simplest explanation is often the best. He champions Bayesian methods, which naturally balance these two ideas through a process called marginalization, which he describes as an automatic Occams razor.
TOC:
[00:00:00] Introduction and Thesis
[00:04:19] Challenging Conventional Wisdom
[00:11:17] The Philosophy of a Scientist-Engineer
[00:16:47] Expressiveness, Overfitting, and Bias
[00:28:15] Understanding, Compression, and Kolmogorov Complexity
[01:05:06] The Surprising Power of Generalization
[01:13:21] The Elegance of Bayesian Inference
[01:33:02] The Geometry of Learning
[01:46:28] Practical Advice and The Future of AI
Prof. Andrew Gordon Wilson:
https://x.com/andrewgwils
https://cims.nyu.edu/~andrewgw/
https://scholar.google.com/citations?user=twWX2LIAAAAJ&hl=en
https://www.youtube.com/watch?v=Aja0kZeWRy4
https://www.youtube.com/watch?v=HEp4TOrkwV4
TRANSCRIPT:
https://app.rescript.info/public/share/H4Io1Y7Rr54MM05FuZgAv4yphoukCfkqokyzSYJwCK8
REFS:
Deep Learning is Not So Mysterious or Different [Andrew Gordon Wilson]
https://arxiv.org/abs/2503.02113
Bayesian Deep Learning and a Probabilistic Perspective of Generalization [Andrew Gordon Wilson, Pavel Izmailov]
https://arxiv.org/abs/2002.08791
Compute-Optimal LLMs Provably Generalize Better With Scale [Marc Finzi, Sanyam Kapoor, Diego Granziol, Anming Gu, Christopher De Sa, J. Zico Kolter, Andrew Gordon Wilson]
https://arxiv.org/abs/2504.15208 The Real Reason Huge AI Models Actually Work [Prof. Andrew Wilson]](https://i.ytimg.com/vi/M-jTeBCEGHc/mqdefault.jpg)
![What If Intelligence Didnt Evolve? It Was There From the Start! - Blaise Agüera y Arcas
Blaise Agüera y Arcas presenting at ALife 2025 — the most technically detailed public walkthrough of the ideas in his *What is Life?* and *What is Intelligence?* books that weve come across.
He covers the BFF experiments (self-replicating programs emerging spontaneously from random noise), the mathematical framework connecting Lotka-Volterra population dynamics with Smoluchowski coagulation, eigenvalue analysis of cooperation matrices, and his central claim that symbiogenesis — not mutation — is the primary engine of evolutionary novelty.
The experimental results are genuinely striking: complex self-replicating code arising from random byte strings with zero mutation, a sharp phase transition that looks like gelation, and a proof that blocking deep symbiogenetic ancestry trees prevents the transition entirely.
A few things worth flagging for critical viewers:
— The substrate is more carefully engineered than the framing sometimes suggests. The choice of language, tape length, interaction protocol, and step limits all shape what emerges. Their own SUBLEQ counterexample (where self-replicators *dont* arise despite being theoretically possible) highlights that these design choices matter substantially — and a general theory of which substrates support this transition is still missing.
— The leap from self-replicating programs on fixed-length tapes to life was computational and intelligent from the start involves significant philosophical extrapolation beyond what the experiments directly demonstrate.
— The Bedau et al. (2000) open problems paper he references at the start actually sets a higher bar for Challenge 3.2 than BFF currently meets: it asks that the internal organization of these organisms and the boundaries separating them from their environment arise and be sustained through the activities of lower-level primitives — whereas BFFs tape boundaries are fixed by design, not emergent.
TIMESTAMPS:
00:00:00 Introduction: From Noise to Programs & ALife History
00:03:15 Defining Life: Function as the Spirit
00:05:45 Von Neumanns Insight: Life is Embodied Computation
00:09:15 Physics of Computation: Irreversibility & Fallacies
00:15:00 The BFF Experiment: Spontaneous Generation of Code
00:23:45 The Mystery: Complexity Growth Without Mutation
00:27:00 Symbiogenesis: The Engine of Novelty
00:33:15 Mathematical Proof: Blocking Symbiosis Stops Life
00:40:15 Evolutionary Implications: Its Symbiogenesis All The Way Down
00:44:30 Intelligence as Modeling Others
00:46:49 Q&A: Levels of Abstraction & Definitions
REFERENCES:
Paper:
[00:01:16] Open Problems in Artificial Life
https://direct.mit.edu/artl/article/6/4/363/2354/Open-Problems-in-Artificial-Life
[00:09:30] When does a physical system compute?
https://arxiv.org/abs/1309.7979
[00:15:00] Computational Life
https://arxiv.org/abs/2406.19108
[00:27:30] On the Origin of Mitosing Cells
https://pubmed.ncbi.nlm.nih.gov/11541392/
[00:42:00] The Major Evolutionary Transitions
https://www.nature.com/articles/374227a0
[00:44:00] The ARC gene
https://www.nih.gov/news-events/news-releases/memory-gene-goes-viral
Person:
[00:05:45] Alan Turing
https://plato.stanford.edu/entries/turing/
[00:07:30] John von Neumann
https://en.wikipedia.org/wiki/John_von_Neumann
[00:11:15] Hector Zenil
https://hectorzenil.net/
[00:12:00] Robert Sapolsky
https://profiles.stanford.edu/robert-sapolsky
[00:29:30] Marian Smoluchowski
https://en.wikipedia.org/wiki/Marian_Smoluchowski
Book:
[00:06:15] What is Life?
https://mitpress.mit.edu/9780262554091/what-is-life/
[00:19:45] What is Life? How Chemistry Becomes Biology
https://amazon.com/dp/0199641013
Technical Concept:
[00:15:45] Brainfuck
https://esolangs.org/wiki/Brainfuck
LINKS:
RESCRIPT: https://app.rescript.info/public/share/ff7gb6HpezOR3DF-gr9-rCoMFzzEgUjLQK6voV5XVWY What If Intelligence Didnt Evolve? It Was There From the Start! - Blaise Agüera y Arcas](https://i.ytimg.com/vi/M2iX6HQOoLg/mqdefault.jpg)
![ARC Prize Version 2 Launch Video! [Francois Chollet, Mike Knoop]
Francois Chollet and Mike Knoop join Tim Scarfe to announce ARC-AGI 2 and the ARC Prize 2025 contest. The new benchmark has been human-calibrated with 400 participants and adversarially designed so that frontier reasoning models score in the single digits, while every task remains solvable by at least two humans. The conversation covers the story behind testing OpenAI o3 on ARC v1 (where it scored 75-85%), what those results reveal about fluid intelligence versus brute-force capability, and why Chollet considers o3 the first model with genuine if limited fluid intelligence. Technical discussion ranges from chain-of-thought reasoning as natural language program synthesis, to the exponential failure modes of current reasoning systems on spatially complex tasks, to the deeper question of whether intelligence is best understood as a binary category or a spectrum of recombination depth. Chollet argues that intelligence is fundamentally about efficiency of knowledge acquisition and adaptation, not raw capability, and that ARC-AGI 2 is designed to measure exactly this distinction.
SPONSOR MESSAGES:
***
Tufa AI Labs is a brand new research lab in Zurich started by Benjamin Crouzier focussed on o-series style reasoning and AGI. They are hiring a Chief Engineer and ML engineers. Events in Zurich.
REFERENCES:
Paper:
[00:00:15] On the Measure of Intelligence
https://arxiv.org/abs/1911.01547
[00:12:50] OpenAI o3 Performance on ARC v1
https://arcprize.org/blog/oai-o3-pub-breakthrough
[00:18:30] Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
https://arxiv.org/abs/2201.11903
[00:26:05] ARC Prize 2024: Technical Report
https://arxiv.org/abs/2412.04604
[00:48:57] The Bitter Lesson
http://www.incompleteideas.net/IncIdeas/BitterLesson.html
[00:53:30] Decoding Strategies in Neural Text Generation
https://www.mdpi.com/2078-2489/12/9/355/pdf
Organization:
[00:06:45] ARC Prize Foundation
https://arcprize.org/
LINKS:
Full Transcript: https://app.rescript.info/share/19ed0e8921636ce804d8780e6b03f67d
Download PDF transcript: https://app.rescript.info/api/public/sessions/302bc045019e056a/pdf ARC Prize Version 2 Launch Video! [Francois Chollet, Mike Knoop]](https://i.ytimg.com/vi/M3b59lZYBW8/mqdefault.jpg)



![You dont fine-tune your way to AGI - Heres why. [Eiso Kant]
Eiso Kant is the co-founder and CTO of Poolside AI, one of roughly seven companies worldwide with the technical muscle to build frontier foundation models from scratch. He sat down with Tim to explain why Poolside deliberately rejected the prevailing wisdom of just scale up the next GPT and instead bet the company on a thesis most labs were ignoring: that reinforcement learning from code execution feedback is the missing scaling axis.
The argument is straightforward and, once you hear it, hard to unsee. Next-token prediction is imitation learning. Reinforcement learning is trial-and-error learning. Poolside maintains close to a million fully containerized code repositories each with its own test suite as a massive, diverse RL environment. The model writes code, executes it, gets deterministic feedback, and learns. This is why software turns out to be the ideal domain for RL scaling: it is deterministic enough to provide clear reward signals, but diverse enough to avoid model collapse.
Along the way, Eiso and Tim get into the weeds on frontier lab operations (over 4000 experimental runs per month), why Chinchilla optimality breaks down once you account for inference cost, what DeepSeek V3 tells us about the second-generation AI company playbook, and why the R1 zero-shot reasoning result should have been the real headline rather than the dollar figure. There is a candid exchange on whether the software development lifecycle itself will progressively collapse into the model, what Chris Olah interpretability work means for alignment, and whether Karpathy was right about Software 2.0 with some important caveats Eiso has developed after a decade of building AI for code.
SPONSOR MESSAGES:
***
Tufa AI Labs is a brand new research lab in Zurich started by Benjamin Crouzier focussed on o-series style reasoning and AGI. They are hiring a Chief Engineer and ML engineers. Events in Zurich.
TIMESTAMPS:
00:00:00 Introduction and Guest Background
00:02:50 Poolside AIs Vision and Three-Step Plan
00:06:50 Foundation Models vs. Enterprise Customization
00:10:25 The Missing Scaling Axis: Reinforcement Learning
00:15:40 Reinforcement Learning from Code Execution Feedback
00:22:20 Model Economics and Experimental Optimization
00:26:00 Enterprise Deployment Strategy and Market Focus
00:30:30 DeepSeek, Distributed Training, and Hardware Architecture
00:36:40 Emergent Reasoning and Chain-of-Thought Scaling
00:45:00 AI-Assisted Software Development Today
00:58:20 Architecture Innovation and Model Interpretability
01:15:00 Karpathys Software 2.0 and the Future of Code
01:25:00 AWS Partnership, Enterprise Security, and Closing
REFERENCES:
website:
[00:01:40] Tufa AI Labs
https://tufalabs.ai/
[00:02:50] Poolside AI
https://poolside.ai/
social:
[00:02:50] Eiso Kant on X
https://x.com/eisokant
paper:
[00:15:40] The Curse of Recursion: Training on Generated Data Makes Models Forget
https://arxiv.org/abs/2305.17493
[00:20:00] Efficient Estimation of Word Representations in Vector Space (Word2Vec)
https://arxiv.org/abs/1301.3781
[00:22:40] On the Measure of Intelligence
https://arxiv.org/abs/1911.01547
[00:30:30] DeepSeek-V3 Technical Report
https://arxiv.org/abs/2412.19437
[00:34:30] Training Compute-Optimal Large Language Models (Chinchilla)
https://arxiv.org/abs/2203.15556
[00:45:45] The Bitter Lesson
http://www.incompleteideas.net/IncIdeas/BitterLesson.html
[00:49:55] Mastering the Game of Go with Deep Neural Networks and Tree Search
https://www.nature.com/articles/nature16961
[01:10:15] Mamba: Linear-Time Sequence Modeling with Selective State Spaces
https://arxiv.org/abs/2312.00752
[01:14:25] Zoom In: An Introduction to Circuits
https://distill.pub/2020/circuits/zoom-in/
benchmark:
[00:46:10] ARC Prize Challenge
https://arcprize.org/
blog:
[01:19:10] Software 2.0
https://karpathy.medium.com/software-2-0-a64152b37c35
LINKS:
Full Transcript: https://app.rescript.info/share/a8144052b0a5fc7210ab37f9651a3557
Download PDF transcript: https://app.rescript.info/api/public/sessions/2f5beb6d70f79973/pdf
Eiso Kant:
https://x.com/eisokant
https://poolside.ai/ You dont fine-tune your way to AGI - Heres why. [Eiso Kant]](https://i.ytimg.com/vi/NDrosuKhXeo/mqdefault.jpg)
![What Building an AI Scientist Actually Requires Beyond Intelligence — Edward Hughes
Can a machine learn the judgement that separates a plausible-looking result from a faithful experiment? Edward Hughes, Chief Scientist and co-founder of Inherent, joins Tim Scarfe to argue that creativity is not optimisation, and that the missing capability in AI is choosing which questions are worth asking.
SPONSOR:
Cyber Fund built the Monastery to help founders ship products that were impossible a year ago.
Apply now: https://cyber.fund
Edward makes the case that Move 37 was innovative rather than creative, and that the field, not the individual, decides what counts as a discovery. That reframing runs through Csikszentmihalyi, Deutsch and exaptation into open-endedness, where deceptive goals and imperfect world models turn out to be the point rather than the problem. The second half turns to the paper: Replica, a task space built by redacting figures from real papers, and Faraday, a 27-billion-parameter model trained to steer a frontier coding agent that then beats the frontier on held-out replications.
TIMESTAMPS:
00:00:00 Cold open: Move 37, Faraday and collective intelligence
00:01:08 Sponsor: CyberFund
00:01:46 Inherents $50M raise and the road from string theory
00:09:14 Three timescales of learning: weights, context, culture
00:13:47 Move 37 was innovative, not creative: the field decides
00:20:39 Creativity as satisficing: the urinal and evolution
00:25:06 Exaptation and the Tristan chord: creativity in context
00:30:56 Coherence for whom? Deutschs hard-to-vary explanations
00:35:53 Why copying is creative: Deutsch and the constraint engineer
00:42:27 Societies of agents and the strong Moravec paradox
00:45:51 Evaluate in hindsight: from Lean proofs to climate change
00:51:56 Picbreeder, local goals and why discovery needs deception
00:57:21 Spaghetti proofs, translation layers and superhuman Go
01:00:37 Does nature compress? Naturalness and real patterns
01:07:36 Why replicate? Replicas redacted figures and Faraday
01:12:31 Faraday beats Codex, Claude and GLM 5.2 on held-out tasks
01:15:31 Replication to innovation: how the Transformer happened
01:18:26 Deep replication: what Faraday learns from Voyager and GNoME
01:23:37 Can the AI scientist cheat? Goodharting the judge
01:29:09 Inside Replica: scale-down, 8xB300 runs, per-task rubrics
01:34:11 The RL crisis: getting GRPO to work with per-turn credit
01:39:43 Weights vs harnesses: AlphaEvolve, DGM and EvoTune
01:45:45 The recursive company: agents cross a phase transition
01:50:35 Collective intelligence and the electric dynamo
01:55:46 What replaces OKRs? Incumbents and the burden of knowledge
REFERENCES:
MLST creativity article:
https://archive.mlst.ai/read/why-creativity-cannot-be-interpolated
organization:
[00:01:47] Inherent
https://inherentlabs.ai/
other:
[00:20:51] Marcel Duchamp, Fountain (1917)
https://www.tate.org.uk/art/artworks/duchamp-fountain-t07573
[00:28:25] The Tristan chord
https://imslp.org/wiki/Tristan_und_Isolde%2C_WWV_90_(Wagner%2C_Richard)
https://arxiv.org/abs/2301.07608
[00:05:41] Genie
https://arxiv.org/abs/2402.15391
[00:06:05] The AI Scientist
https://arxiv.org/abs/2408.06292
[00:12:13] Training AI Scientists to Replicate Research (Replica and Faraday)
https://arxiv.org/abs/2608.13331
[00:25:42] Open-Endedness is Essential for ASI
https://arxiv.org/abs/2406.04268
[00:50:30] MAP-Elites
https://arxiv.org/abs/1504.04909
[00:50:59] OMNI: Open-endedness via Models of human Notions of Interestingness
https://arxiv.org/abs/2306.01711
[00:51:01] OMNI-EPIC
https://arxiv.org/abs/2405.15568
[00:53:42] Picbreeder
https://pubmed.ncbi.nlm.nih.gov/20964537/
[01:17:26] Neural Machine Translation by Jointly Learning to Align and Translate
https://arxiv.org/abs/1409.0473
[01:20:44] Voyager
https://arxiv.org/abs/2305.16291
[01:21:57] GNoME
https://www.nature.com/articles/s41586-023-06735-9
[01:34:11] DeepSeekMath
https://arxiv.org/abs/2402.03300
[01:41:29] AlphaEvolve
https://arxiv.org/abs/2506.13131
[01:41:53] Darwin Gödel Machine
https://arxiv.org/abs/2505.22954
[01:41:56] Hyperagents
https://arxiv.org/abs/2603.19461
[01:42:33] EvoTune
https://arxiv.org/abs/2504.05108
[01:44:46] Evolutionary Principles in Self-Referential Learning
https://people.idsia.ch/~juergen/diploma.html
[01:59:33] Are Ideas Getting Harder to Find?
https://www.nber.org/papers/w23782
book:
[00:16:04] Creativity: Flow and the Psychology of Discovery and Invention
https://search.worldcat.org/title/254487436
[00:26:22] Why Greatness Cannot Be Planned
https://link.springer.com/book/10.1007/978-3-319-15524-1
[00:33:03] The Beginning of Infinity
https://www.penguinrandomhouse.com/books/293575/the-beginning-of-infinity-by-david-deutsch/
[01:55:47] Laws of Knowledge
https://www.penguin.co.nz/books/the-infinite-alphabet-9780241655672
RESCRIPT:
https://app.rescript.info/session/670296ba913761d0?share=6281911cac9bdbff637f10819d4d1e5c What Building an AI Scientist Actually Requires Beyond Intelligence — Edward Hughes](https://i.ytimg.com/vi/P4bYjTJvD28/mqdefault.jpg)
![Can Latent Program Networks Solve Abstract Reasoning? [Clement Bonnet]
SPONSOR MESSAGES:
***
CentML offers competitive pricing for GenAI model deployment, with flexible options to suit a wide range of models, from small to large-scale deployments. Check out their super fast DeepSeek R1 hosting!
https://centml.ai/pricing/
Clement Bonnet presents Latent Program Networks (LPNs), a fundamentally different approach to the ARC-AGI benchmark that embeds programs into a continuous latent space rather than generating code or fine-tuning language models. The architecture uses a VAE-style encoder to map input-output pairs into distributional program representations, performs gradient-based search during both training and inference to refine these representations, and decodes solutions directly without an explicit program intermediary.
The conversation explores why ARC remains impervious to standard neural networks Chollet designed it to resist memorization, with test tasks far enough from any training distribution that zero-shot generalization fails. Bonnet and Tim Scarfe dig into the induction vs. transduction distinction, the role of the kernel trick in blurring that boundary, and why parameter-space search (test-time training) may be less efficient than searching a compressed program manifold.
A key architectural insight: training the search procedure end-to-end forces the latent space to become smooth and searchable, similar to MAML-style meta-learning. The system achieves ~10% on ARC evaluation sets using only small transformers (40M parameters total) trained from scratch on Michael Hodels re-ARC dataset no pre-trained LLMs, no internet priors.
The discussion also probes the limits of continuous latent spaces for compositional reasoning. Bonnet acknowledges that single-thread search through a fixed latent space cannot achieve true program composition, but suggests multi-threaded search with external composition could bridge the gap. The episode closes with reflections on creativity, the efficiency gap between human hypothesis-testing (a handful of guesses) and LLM program sampling (millions of samples), and what scaling LPNs might reveal about the structure of program spaces.
REFERENCES:
paper:
[00:00:05] ARC-AGI Benchmark
https://arxiv.org/abs/2412.04604
[00:02:10] Latent Program Networks
https://arxiv.org/abs/2411.08706
[00:08:45] Induction vs Transduction in Abstract Reasoning
https://arxiv.org/abs/2411.02272
[00:17:40] Variational Autoencoders
https://arxiv.org/abs/1312.6114
[00:33:00] Critique of Deep Learning
https://arxiv.org/abs/2002.06177
person:
[00:07:45] Kevin Ellis - Program Synthesis
https://www.cs.cornell.edu/~ellisk/
tool:
[00:27:50] re-ARC Framework
https://github.com/michaelhodel/re-arc
LINKS:
Full Transcript: https://app.rescript.info/share/55768bb71fbcc62c0522136650ccf960
Download PDF transcript: https://app.rescript.info/api/public/sessions/54ec5e9b3152760b/pdf
Clem and Matthew-
https://www.linkedin.com/in/clement-bonnet16/
https://github.com/clement-bonnet
https://mvmacfarlane.github.io/
[00:27:50] re-ARC, Hodel
https://github.com/michaelhodel/re-arc
[00:29:40] Grid size in ARC tasks, Chollet
https://github.com/fchollet/ARC-AGI Can Latent Program Networks Solve Abstract Reasoning? [Clement Bonnet]](https://i.ytimg.com/vi/PHBItVuudbU/mqdefault.jpg)