Machine Learning Street Talk (MLST)

We often think of Large Language Models (LLMs) as all-knowing, but as the team reveals, they still struggle with the logic of a second-grader. Why can’t ChatGPT reliably add large numbers? Why does it "hallucinate" the laws of physics? The answer lies in the architecture. This episode explores how *Category Theory* —an ultra-abstract branch of mathematics—could provide the "Periodic Table" for neural networks, turning the "alchemy" of modern AI into a rigorous science.In this deep-dive exploration, *Andrew Dudzik*, *Petar Velichkovich*, *Taco Cohen*, *Bruno Gavranović*, and *Paul Lessard* join host *Tim Scarfe* to discuss the fundamental limitations of today’s AI and the radical mathematical framework that might fix them.TRANSCRIPT:https://app.rescript.info/public/share/LMreunA-BUpgP-2AkuEvxA7BAFuA-VJNAp2Ut4MkMWk---Key Insights in This Episode:* *The "Addition" Problem:* *Andrew Dudzik* explains why LLMs don't actually "know" math—they just recognize patterns. When you change a single digit in a long string of numbers, the pattern breaks because the model lacks the internal "machinery" to perform a simple carry operation.* *Beyond Alchemy:* deep learning is currently in its "alchemy" phase—we have powerful results, but we lack a unifying theory. Category Theory is proposed as the framework to move AI from trial-and-error to principled engineering. [00:13:49]* *Algebra with Colors:* To make Category Theory accessible, the guests use brilliant analogies—like thinking of matrices as *magnets with colors* that only snap together when the types match. This "partial compositionality" is the secret to building more complex internal reasoning. [00:09:17]* *Synthetic vs. Analytic Math:* *Paul Lessard* breaks down the philosophical shift needed in AI research: moving from "Analytic" math (what things are made of) to "Synthetic" math [00:23:41]---Why This Matters for AGIIf we want AI to solve the world's hardest scientific problems, it can't just be a "stochastic parrot." It needs to internalize the rules of logic and computation. By imbuing neural networks with categorical priors, researchers are attempting to build a future where AI doesn't just predict the next word—it understands the underlying structure of the universe.---TIMESTAMPS:00:00:00 The Failure of LLM Addition & Physics00:01:26 Tool Use vs Intrinsic Model Quality00:03:07 Efficiency Gains via Internalization00:04:28 Geometric Deep Learning & Equivariance00:07:05 Limitations of Group Theory00:09:17 Category Theory: Algebra with Colors00:11:25 The Systematic Guide of Lego-like Math00:13:49 The Alchemy Analogy & Unifying Theory00:15:33 Information Destruction & Reasoning00:18:00 Pathfinding & Monoids in Computation00:20:15 System 2 Reasoning & Error Awareness00:23:31 Analytic vs Synthetic Mathematics00:25:52 Morphisms & Weight Tying Basics00:26:48 2-Categories & Weight Sharing Theory00:28:55 Higher Categories & Emergence00:31:41 Compositionality & Recursive Folds00:34:05 Syntax vs Semantics in Network Design00:36:14 Homomorphisms & Multi-Sorted Syntax00:39:30 The Carrying Problem & Hopf FibrationsPetar Veličković (GDM)https://petar-v.com/Paul Lessardhttps://www.linkedin.com/in/paul-roy-lessard/Bruno Gavranovićhttps://www.brunogavranovic.com/Andrew Dudzik (GDM)https://www.linkedin.com/in/andrew-dudzik-222789142/---REFERENCES:Model:[00:01:05] Veohttps://deepmind.google/models/veo/[00:01:10] Geniehttps://deepmind.google/blog/genie-3-a-new-frontier-for-world-models/Paper:[00:04:30] Geometric Deep Learning Blueprinthttps://arxiv.org/abs/2104.13478https://www.youtube.com/watch?v=bIZB1hIJ4u8[00:16:45] AlphaGeometryhttps://arxiv.org/abs/2401.08312[00:16:55] AlphaCodehttps://arxiv.org/abs/2203.07814[00:17:05] FunSearchhttps://www.nature.com/articles/s41586-023-06924-6[00:37:00] Attention Is All You Needhttps://arxiv.org/abs/1706.03762[00:43:00] Categorical Deep Learninghttps://arxiv.org/abs/2402.15332

What if everything we think we know about AI understanding is wrong? Is compression the key to intelligence? Or is there something more—a leap from memorization to true abstraction? In this fascinating conversation, we sit down with **Professor Yi Ma**—world-renowned expert in deep learning, IEEE/ACM Fellow, and author of the groundbreaking new book *Learning Deep Representations of Data Distributions*. Professor Ma challenges our assumptions about what large language models actually do, reveals why 3D reconstruction isn't the same as understanding, and presents a unified mathematical theory of intelligence built on just two principles: **parsimony** and **self-consistency**.**SPONSOR MESSAGES START**—Prolific - Quality data. From real people. For faster breakthroughs.https://www.prolific.com/?utm_source=mlst—cyber•Fund https://cyber.fund/?utm_source=mlst is a founder-led investment firm accelerating the cybernetic economyHiring a SF VC Principal: https://talent.cyber.fund/companies/cyber-fund-2/jobs/57674170-ai-investment-principal#content?utm_source=mlstSubmit investment deck: https://cyber.fund/contact?utm_source=mlst—**END**Key Insights:**LLMs Don't Understand—They Memorize**Language models process text (*already* compressed human knowledge) using the same mechanism we use to learn from raw data. **The Illusion of 3D Vision**Sora and NeRFs etc that can reconstruct 3D scenes still fail miserably at basic spatial reasoning**"All Roads Lead to Rome"**Why adding noise is *necessary* for discovering structure.**Why Gradient Descent Actually Works**Natural optimization landscapes are surprisingly smooth—a "blessing of dimensionality" **Transformers from First Principles**Transformer architectures can be mathematically derived from compression principles—INTERACTIVE AI TRANSCRIPT PLAYER w/REFS (ReScript):https://app.rescript.info/public/share/Z-dMPiUhXaeMEcdeU6Bz84GOVsvdcfxU_8Ptu6CTKMQAbout Professor Yi MaYi Ma is the inaugural director of the School of Computing and Data Science at Hong Kong University and a visiting professor at UC Berkeley. https://people.eecs.berkeley.edu/~yima/https://scholar.google.com/citations?user=XqLiBQMAAAAJ&hl=en https://x.com/YiMaTweets **Slides from this conversation:**https://www.dropbox.com/scl/fi/sbhbyievw7idup8j06mlr/slides.pdf?rlkey=7ptovemezo8bj8tkhfi393fh9&dl=0**Related Talks by Professor Ma:**- Pursuing the Nature of Intelligence (ICLR): https://www.youtube.com/watch?v=LT-F0xSNSjo- Earlier talk at Berkeley: https://www.youtube.com/watch?v=TihaCUjyRLMTIMESTAMPS:00:00:00 Introduction00:02:08 The First Principles Book & Research Vision00:05:21 Two Pillars: Parsimony & Consistency00:09:50 Evolution vs. Learning: The Compression Mechanism00:14:36 LLMs: Memorization Masquerading as Understanding00:19:55 The Leap to Abstraction: Empirical vs. Scientific00:27:30 Platonism, Deduction & The ARC Challenge00:35:57 Specialization & The Cybernetic Legacy00:41:23 Deriving Maximum Rate Reduction00:48:21 The Illusion of 3D Understanding: Sora & NeRF00:54:26 All Roads Lead to Rome: The Role of Noise00:59:56 All Roads Lead to Rome: The Role of Noise01:00:14 Benign Non-Convexity: Why Optimization Works01:06:35 Double Descent & The Myth of Overfitting01:14:26 Self-Consistency: Closed-Loop Learning01:21:03 Deriving Transformers from First Principles01:30:11 Verification & The Kevin Murphy Question01:34:11 CRATE vs. ViT: White-Box AI & ConclusionREFERENCES:Book:[00:03:04] Learning Deep Representations of Data Distributionshttps://ma-lab-berkeley.github.io/deep-representation-learning-book/[00:18:38] A Brief History of Intelligencehttps://www.amazon.co.uk/BRIEF-HISTORY-INTELLIGEN-HB-Evolution/dp/0008560099[00:38:14] Cyberneticshttps://mitpress.mit.edu/9780262730099/cybernetics/Book (Yi Ma):[00:03:14] 3-D Vision bookhttps://link.springer.com/book/10.1007/978-0-387-21779-6<TRUNC> refs on ReScript link/YT

The Transformer architecture (which powers ChatGPT and nearly all modern AI) might be trapping the industry in a localized rut, preventing us from finding true intelligent reasoning, according to the person who co-invented it. Llion Jones and Luke Darlow, key figures at the research lab Sakana AI, join the show to make this provocative argument, and also introduce new research which might lead the way forwards.**SPONSOR MESSAGES START**—Build your ideas with AI Studio from Google - http://ai.studio/build—Tufa AI Labs is hiring ML Research Engineers https://tufalabs.ai/ —cyber•Fund https://cyber.fund/?utm_source=mlst is a founder-led investment firm accelerating the cybernetic economyHiring a SF VC Principal: https://talent.cyber.fund/companies/cyber-fund-2/jobs/57674170-ai-investment-principal#content?utm_source=mlstSubmit investment deck: https://cyber.fund/contact?utm_source=mlst—**END**The "Spiral" Problem – Llion uses a striking visual analogy to explain what current AI is missing. If you ask a standard neural network to understand a spiral shape, it solves it by drawing tiny straight lines that just happen to look like a spiral. It "fakes" the shape without understanding the concept of spiraling. Introducing the Continuous Thought Machine (CTM) Luke Darlow deep dives into their solution: a biology-inspired model that fundamentally changes how AI processes information.The Maze Analogy: Luke explains that standard AI tries to solve a maze by staring at the whole image and guessing the entire path instantly. Their new machine "walks" through the maze step-by-step.Thinking Time: This allows the AI to "ponder." If a problem is hard, the model can naturally spend more time thinking about it before answering, effectively allowing it to correct its own mistakes and backtrack—something current Language Models struggle to do genuinely.https://sakana.ai/https://x.com/YesThisIsLionhttps://x.com/LearningLukeDTRANSCRIPT:https://app.rescript.info/public/share/crjzQ-Jo2FQsJc97xsBdfzfOIeMONpg0TFBuCgV2Fu8TOC:00:00:00 - Stepping Back from Transformers00:00:43 - Introduction to Continuous Thought Machines (CTM)00:01:09 - The Changing Atmosphere of AI Research00:04:13 - Sakana’s Philosophy: Research Freedom00:07:45 - The Local Minimum of Large Language Models00:18:30 - Representation Problems: The Spiral Example00:29:12 - Technical Deep Dive: CTM Architecture00:36:00 - Adaptive Computation & Maze Solving00:47:15 - Model Calibration & Uncertainty01:00:43 - Sudoku Bench: Measuring True ReasoningREFS:Why Greatness Cannot be planned [Kenneth Stanley]https://www.amazon.co.uk/Why-Greatness-Cannot-Planned-Objective/dp/3319155237https://www.youtube.com/watch?v=lhYGXYeMq_E The Hardware Lottery [Sara Hooker]https://arxiv.org/abs/2009.06489https://www.youtube.com/watch?v=sQFxbQ7ade0 Continuous Thought Machines [Luke Darlow et al / Sakana]https://arxiv.org/abs/2505.05522https://sakana.ai/ctm/ LSTM: The Comeback Story? [Prof. Sepp Hochreiter]https://www.youtube.com/watch?v=8u2pW2zZLCs Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis [Kumar/Stanley]https://arxiv.org/pdf/2505.11581 A Spline Theory of Deep Networks [Randall Balestriero]https://proceedings.mlr.press/v80/balestriero18b/balestriero18b.pdf https://www.youtube.com/watch?v=86ib0sfdFtw https://www.youtube.com/watch?v=l3O2J3LMxqI On the Biology of a Large Language Model [Anthropic, Jack Lindsey et al]https://transformer-circuits.pub/2025/attribution-graphs/biology.html The ARC Prize 2024 Winning Algorithm [Daniel Franzen and Jan Disselhoff] “The ARChitects”https://www.youtube.com/watch?v=mTX_sAq--zYNeural Turing Machine [Graves]https://arxiv.org/pdf/1410.5401 Adaptive Computation Time for Recurrent Neural Networks [Graves]https://arxiv.org/abs/1603.08983 Sudoko Bench [Sakana] https://pub.sakana.ai/sudoku/

"What is life?" - asks Chris Kempes, a professor at the Santa Fe Institute.Chris explains that scientists are moving beyond a purely Earth-based, biological view and are searching for a universal theory of life that could apply to anything, anywhere in the universe. He proposes that things we don't normally consider "alive"—like human culture, language, or even artificial intelligence; could be seen as life forms existing on different "substrates".To understand this, Chris presents a fascinating three-level framework:- Materials: The physical stuff life is made of. He argues this could be incredibly diverse across the universe, and we shouldn't expect alien life to share our biochemistry.- Constraints: The universal laws of physics (like gravity or diffusion) that all life must obey, regardless of what it's made of. This is where different life forms start to look more similar.- Principles: At the highest level are abstract principles like evolution and learning. Chris suggests these computational or "optimization" rules are what truly define a living system.A key idea is "convergence" – using the example of the eye. It's such a complex organ that you'd think it evolved only once. However, eyes evolved many separate times across different species. This is because the physics of light provides a clear "target", and evolution found similar solutions to the problem of seeing, even with different starting materials.**SPONSOR MESSAGES**—Prolific - Quality data. From real people. For faster breakthroughs.https://www.prolific.com/?utm_source=mlst—Check out NotebookLM from Google here - https://notebooklm.google.com/ - it’s really good for doing research directly from authoritative source material, minimising hallucinations. —cyber•Fund https://cyber.fund/?utm_source=mlst is a founder-led investment firm accelerating the cybernetic economyHiring a SF VC Principal: https://talent.cyber.fund/companies/cyber-fund-2/jobs/57674170-ai-investment-principal#content?utm_source=mlstSubmit investment deck: https://cyber.fund/contact?utm_source=mlst— Prof. Chris Kempes:https://www.santafe.edu/people/profile/chris-kempesTRANSCRIPT:https://app.rescript.info/public/share/Y2cI1i0nX_-iuZitvlguHvaVLQTwPX1Y_E1EHxV0i9ITOC:00:00:00 - Introduction to Chris Kempes and the Santa Fe Institute00:02:28 - The Three Cultures of Science00:05:08 - What Makes a Good Scientific Theory?00:06:50 - The Universal Theory of Life00:09:40 - The Role of Material in Life00:12:50 - A Hierarchy for Understanding Life00:13:55 - How Life Diversifies and Converges00:17:53 - Adaptive Processes and Defining Life00:19:28 - Functionalism, Memes, and Phylogenies00:22:58 - Convergence at Multiple Levels00:25:45 - The Possibility of Simulating Life00:28:16 - Intelligence, Parasitism, and Spectrums of Life00:32:39 - Phase Changes in Evolution00:36:16 - The Separation of Matter and Logic00:37:21 - Assembly Theory and Quantifying ComplexityREFS:Developing a predictive science of the biosphere requires the integration of scientific cultures [Kempes et al]https://www.pnas.org/doi/10.1073/pnas.2209196121Seeing with an extra sense (“Dangerous prediction”) [Rob Phillips]https://www.sciencedirect.com/science/article/pii/S0960982224009035 The Multiple Paths to Multiple Life [Christopher P. Kempes & David C. Krakauer]https://link.springer.com/article/10.1007/s00239-021-10016-2 The Information Theory of Individuality [David Krakauer et al]https://arxiv.org/abs/1412.2447Minds, Brains and Programs [Searle]https://home.csulb.edu/~cwallis/382/readings/482/searle.minds.brains.programs.bbs.1980.pdf The error thresholdhttps://www.sciencedirect.com/science/article/abs/pii/S0168170204003843Assembly theory and its relationship with computational complexity [Kempes et al]https://arxiv.org/abs/2406.12176

We sat down with Sara Saab (VP of Product at Prolific) and Enzo Blindow (VP of Data and AI at Prolific) to explore the critical role of human evaluation in AI development and the challenges of aligning AI systems with human values. Prolific is a human annotation and orchestration platform for AI used by many of the major AI labs. This is a sponsored show in partnership with Prolific. **SPONSOR MESSAGES**—cyber•Fund https://cyber.fund/?utm_source=mlst is a founder-led investment firm accelerating the cybernetic economyOct 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=mlstSubmit investment deck: https://cyber.fund/contact?utm_source=mlst— While technologists want to remove humans from the loop for speed and efficiency, these non-deterministic AI systems actually require more human oversight than ever before. Prolific's approach is to put "well-treated, verified, diversely demographic humans behind an API" - making human feedback as accessible as any other infrastructure service.When AI models like Grok 4 achieve top scores on technical benchmarks but feel awkward or problematic to use in practice, it exposes the limitations of our current evaluation methods. The guests argue that optimizing for benchmarks may actually weaken model performance in other crucial areas, like cultural sensitivity or natural conversation.We also discuss Anthropic's research showing that frontier AI models, when given goals and access to information, independently arrived at solutions involving blackmail - without any prompting toward unethical behavior. Even more concerning, the more sophisticated the model, the more susceptible it was to this "agentic misalignment." Enzo and Sarah present Prolific's "Humane" leaderboard as an alternative to existing benchmarking systems. By stratifying evaluations across diverse demographic groups, they reveal that different populations have vastly different experiences with the same AI models. Looking ahead, the guests imagine a world where humans take on coaching and teaching roles for AI systems - similar to how we might correct a child or review code. This also raises important questions about working conditions and the evolution of labor in an AI-augmented world. Rather than replacing humans entirely, we may be moving toward more sophisticated forms of human-AI collaboration.As AI tech becomes more powerful and general-purpose, the quality of human evaluation becomes more critical, not less. We need more representative evaluation frameworks that capture the messy reality of human values and cultural diversity. Visit Prolific: https://www.prolific.com/Sara Saab (VP Product):https://uk.linkedin.com/in/sarasaabEnzo Blindow (VP Data & AI):https://uk.linkedin.com/in/enzoblindowTRANSCRIPT:https://app.rescript.info/public/share/xZ31-0kJJ_xp4zFSC-bunC8-hJNkHpbm7Lg88RFcuLETOC:[00:00:00] Intro & Background[00:03:16] Human-in-the-Loop Challenges[00:17:19] Can AIs Understand?[00:32:02] Benchmarking & Vibes[00:51:00] Agentic Misalignment Study[01:03:00] Data Quality vs Quantity[01:16:00] Future of AI OversightREFS:Anthropic Agentic Misalignmenthttps://www.anthropic.com/research/agentic-misalignmentValue Compasshttps://arxiv.org/pdf/2409.09586Reasoning Models Don’t Always Say What They Think (Anthropic)https://www.anthropic.com/research/reasoning-models-dont-say-think https://assets.anthropic.com/m/71876fabef0f0ed4/original/reasoning_models_paper.pdfApollo research - science of evals blog posthttps://www.apolloresearch.ai/blog/we-need-a-science-of-evals Leaderboard Illusion https://www.youtube.com/watch?v=9W_OhS38rIE MLST videoThe Leaderboard Illusion [2025]Shivalika Singh et alhttps://arxiv.org/abs/2504.20879(Truncated, full list on YT)

We need AI systems to synthesise new knowledge, not just compress the data they see. Jeremy Berman, is a research scientist at Reflection AI and recent winner of the ARC-AGI v2 public leaderboard.**SPONSOR MESSAGES**—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 economyOct 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=mlstSubmit investment deck: https://cyber.fund/contact?utm_source=mlst— Imagine trying to teach an AI to think like a human i.e. solving puzzles that are easy for us but stump even the smartest models. Jeremy's evolutionary approach—evolving natural language descriptions instead of python code like his last version—landed him at the top with about 30% accuracy on the ARCv2.We discuss why current AIs are like "stochastic parrots" that memorize but struggle to truly reason or innovate as well as big ideas like building "knowledge trees" for real understanding, the limits of neural networks versus symbolic systems, and whether we can train models to synthesize new ideas without forgetting everything else. Jeremy Berman:https://x.com/jerber888TRANSCRIPT:https://app.rescript.info/public/share/qvCioZeZJ4Q_NlR66m-hNUZnh-qWlUJcS15Wc2OGwD0TOC:Introduction and Overview [00:00:00]ARC v1 Solution [00:07:20]Evolutionary Python Approach [00:08:00]Trade-offs in Depth vs. Breadth [00:10:33]ARC v2 Improvements [00:11:45]Natural Language Shift [00:12:35]Model Thinking Enhancements [00:13:05]Neural Networks vs. Symbolism Debate [00:14:24]Turing Completeness Discussion [00:15:24]Continual Learning Challenges [00:19:12]Reasoning and Intelligence [00:29:33]Knowledge Trees and Synthesis [00:50:15]Creativity and Invention [00:56:41]Future Directions and Closing [01:02:30]REFS:Jeremy’s 2024 article on winning ARCAGI1-pubhttps://jeremyberman.substack.com/p/how-i-got-a-record-536-on-arc-agiGetting 50% (SoTA) on ARC-AGI with GPT-4o [Greenblatt]https://blog.redwoodresearch.org/p/getting-50-sota-on-arc-agi-with-gpt https://www.youtube.com/watch?v=z9j3wB1RRGA [his MLST interview]A Thousand Brains: A New Theory of Intelligence [Hawkins]https://www.amazon.com/Thousand-Brains-New-Theory-Intelligence/dp/1541675819https://www.youtube.com/watch?v=6VQILbDqaI4 [MLST interview]Francois Chollet + Mike Knoop’s labhttps://ndea.com/On the Measure of Intelligence [Chollet]https://arxiv.org/abs/1911.01547On the Biology of a Large Language Model [Anthropic]https://transformer-circuits.pub/2025/attribution-graphs/biology.html The ARChitects [won 2024 ARC-AGI-1-private]https://www.youtube.com/watch?v=mTX_sAq--zY Connectionism critique 1998 [Fodor/Pylshyn]https://uh.edu/~garson/F&P1.PDF Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis [Kumar/Stanley]https://arxiv.org/pdf/2505.11581 AlphaEvolve interview (also program synthesis)https://www.youtube.com/watch?v=vC9nAosXrJw ShinkaEvolve: Evolving New Algorithms with LLMs, Orders of Magnitude More Efficiently [Lange et al]https://sakana.ai/shinka-evolve/ Deep learning with Python Rev 3 [Chollet] - READ CHAPTER 19 NOW!https://deeplearningwithpython.io/