1 hour left before this week’s meetup of our Generative AI study group starts! If you haven’t registered yet, head over now to and don’t forget to join the #generative-ai# Slack channel. See you there!

Visit to register for tomorrow's Generative AI study group at 8 am PT. Discuss distributed inference pipelines, exchange ideas on reasoning-time scaling techniques, explore multimodal transformer models, and more.

Join us for another engaging conversation around open-source distributed inference frameworks, LVLM spatial understanding, AI observability for agentic systems, and more in our Generative AI study group! Visit to register and we’ll see you tomorrow at 8 am PT.

Join our weekly Generative AI study group every Friday at 8 am PT! Dig into geospatial reasoning, share strategies for patch-text alignment, discuss KV cache routing, and more! For registration, head over to to sign up.

As reasoning models consume more tokens and AI systems become more expensive to run, understanding what those tokens actually buy is becoming increasingly important. In this episode, @Stanford professor and Big Spin co-founder @ChrisGPotts joins us to discuss AI tokenomics and his research into “tokenflation”—the possibility that token usage is growing faster than the measurable value those tokens produce. We explore how to measure the return on AI spending, why benchmarks alone provide an incomplete picture of model progress, and what inference-time scaling means for the economics of increasingly capable models. Chris also explains why expert AI users tend to get better results by challenging and iterating with models, how AI fluency affects outcomes, and why more efficient architectures could change the underlying economics. We also discuss DSPy, interpretability, the limits of today’s transformer architectures, and where Chris sees opportunities for more fundamental innovation in AI. 🗒️ Full show notes: 📖 CHAPTERS =============================== 00:00 - Introduction 05:38 - Linguistics in the Age of Language Models 09:26 - Scale Limitations in NLP Research 12:54 - Challenging the Bitter Lesson Mindset 15:12 - Relationship Between Data, Mechanistic Interpretability, and Efficiency 17:03 - DSPy 21:32 - Prompt Optimization and Model Variability 24:35 - Tokenomics and the Rising Cost of AI 28:13 - Measuring Token Purchasing Power with a CPI 32:18 - Inference-Time Scaling 35:36 - Defining Value Across Different AI Tasks 38:33 - AI Value Creation 40:38 - Predicting AI Costs 42:25 - Tokenflation 46:22 - AI Fluency 50:12 - Key Lessons of AI Fluency Work 54:44 - Future Directions

Be part of tomorrow's discussion in our Generative AI study group and dig into long-horizon autonomous agents, explore prompt caching techniques, discuss sparse grouped-query attention (GQA), and more! The meetup starts at 8 am PT. Register today at and see you there!

Our Generative AI study group is open to everyone! Just register at and join us every Friday at 8 am PT. Share RAG grounding techniques, discuss long-context transformer models, explore Cache-Augmented Generation (CAG), and more! We’re looking forward to you joining us!

Secure your spot for tomorrow's session of our Generative AI study group by registering at Discuss vision-language agent workflows, share effective pruning techniques, explore graph neural networks, and more! We’ll meet you there tomorrow at 8 am PT.

Be part of our weekly Generative AI study group every Friday at 8 am PT. Exchange model merging techniques, discuss TPU orchestration, dig into interpretability frameworks, and more! Make sure to register at and join the #generative-ai# Slack channel. See you there!

The conventional wisdom in AI is that the next breakthrough will come from more compute, more data, and larger models. But what if the next leap comes from somewhere else? In this episode, @wellingmax—co-founder and CTO of @cusp_ai and professor at the University of Amsterdam—argues that physics may provide some of the ideas behind the next generation of AI systems. We begin with CuspAI’s work using generative AI to design entirely new materials for semiconductors, batteries, carbon capture, and clean energy. Max explains how foundation models for chemistry, agentic workflows, simulation, and automated experimentation are dramatically accelerating the search for new materials and reshaping scientific discovery. The conversation then broadens into a deeper question. Beyond giving AI new scientific problems to solve, can physics also teach us how to build better AI? Max explores surprising connections between machine learning and thermodynamics, why waves may become a new computational primitive for neural networks, and how concepts like symmetry breaking and statistical physics could inspire AI architectures beyond today’s scaling paradigm. 🗒️ Full show notes: 📖 CHAPTERS =============================== 00:00 - Introduction 02:08 - From Equivariant Networks to AI for Science 04:55 - Founding CuspAI 07:30 - Using AI to Discover New Materials 09:03 - Designing Materials for Carbon Capture 13:20 - Partnerships and the CuspAI Business Model 14:59 - The End-to-End Materials Discovery Process 19:13 - Experiments and Self-Driving Labs 23:04 - Open-Source Molecular Dynamics on GPUs 25:23 - Foundation Models for Chemistry 30:20 - The Future of AI-Driven Materials Science 32:22 - Connecting Generative AI and Thermodynamics 39:29 - How Physics and Machine Learning Can Inform Each Other 44:53 - Waves as a New Primitive for Neural Networks 49:34 - Memory, Stability, and the Edge of Chaos 52:40 - Spontaneous Symmetry Breaking in Neural Networks 56:10 - The Two-Way Exchange Between AI and Physics

Be part of tomorrow’s Generative AI study group discussion! Dig into Mixture-of-Experts routing, explore agent coding benchmarks, discuss long-context inference techniques, and more! Sign up at and we’ll catch you there at 8 am PT.

Head over to to register and connect with the community in our Generative AI study group every Friday at 8 am PT. Explore short-context models, share insights on retrieval-augmented reasoning, dig into multimodal agent systems, and more! Meet you there!

Fatih Porikli, VP of Technology at @QCOMResearch, highlights the following CVPR papers: 🪩 DisCo - a reinforcement learning approach that uses Group Relative Policy Optimization (GRPO) to improve identity diversity in generated images using rewards such as intra-image diversity, inter-image diversity, count accuracy, and image quality 🎨 Ar2Can - a novel "architect and artist" framework that separates scene planning and composition from image rendering, similar to how a human artist might work, to produce more controllable and coherent generations. 🖼️ PixelRush tackles the challenge of generating high-resolution images efficiently on mobile devices using cascade upsampling, latent-space refinement, patchification, and semantically guided noise injection to produce high-quality images without visible artifacts. 🖌️ InverFill tackles the challenge of image inpainting using inverted, semantically steered noise to preserve the background and eliminate boundary artifacts 🎥 Dig into the papers here:

Today, we're joined by @achowdhery, member of technical staff at @reflection_ai, to explore the fundamental shifts required to build true agentic AI. While the industry has largely focused on post-training techniques to improve reasoning, Aakanksha draws on her experience leading pre-training efforts for Google’s PaLM and early Gemini models to argue that pre-training itself must be rethought to move beyond static benchmarks. We explore the limitations of next-token prediction for multi-step workflows and examine how attention mechanisms, loss objectives, and training data must evolve to support long-form reasoning and planning. Aakanksha shares insights on the difference between context retrieval and actual reasoning, the importance of "trajectory" training data, and why scaling remains essential for discovering emergent agentic capabilities like error recovery and dynamic tool learning. 🗒️ For the full list of resources for this episode, visit the show notes page: 📖 CHAPTERS =============================== 00:00 - Introduction 02:26 - Reflection 04:54 - Limitations of post-training for building agents 07:31 - Rethinking pre-training in agents 10:51 - Scaling 11:27 - Evolving attention mechanisms for agentic capabilities 12:39 - Memory as a tool 14:13 - Loss objectives and training data 15:50 - Fine-tuning loss in agent performance 19:37 - Training data 21:29 - Augmenting dominant training data source 24:11 - Overcoming challenges in training on synthetic data 25:47 - Benchmarks 30:44 - Scaling laws in large models versus small models 33:20 - Long-form versus short-form reasoning 37:57 - Agent’s ability to recover from failure 40:15 - Hallucinations and failure recovery 43:53 - Tool use in agents 46:38 - Coding agents 48:37 - How researchers can contribute to agentic AI