Oct 10, 2026

From 15% to 90% GPU Utilization: Fix the Data Pipeline, Not the Model

Tarun Sunkaraneni from Amazon AGI shows how a multimodal SFT pipeline that spent about 85% of its time waiting on data was brought to roughly 90% GPU utilization. He walks through concurrency with asyncio and Ray actors, prefetching, zero-copy transfer via Ray's object store, and a network-card bottleneck that appears only at scale, and argues that the hardest part of training throughput is often keeping GPUs fed rather than optimizing the model.

Oct 10, 2026

Why AlphaFold Didn't Solve Protein Folding — Pushmeet Kohli, Google DeepMind & Sal Candido, Biohub

A panel with Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido, moderated by Brandon Anderson, examines how AI should be built for biology. Topics include scaling laws and data quality, the bitter lesson applied to data, AlphaFold's handcrafted design and its limits, protein dynamics, cryo-EM as a richer data source, trust and calibration versus interpretability, and why curing disease requires 10x breakthroughs.

Oct 10, 2026

Small Models, Big Results: Training a Finance Agent for Under $500 — Charles Dickens, Snorkel AI

Snorkel AI and UC Berkeley's Sky Computing Lab show that a 4B Qwen3 model, trained with reinforcement learning for under $500 using the rLLM framework, beats its 235B sibling on expert-validated financial question answering built from SEC 10-K filings. The talk covers the FinQA dataset construction and verification, common failure modes of frontier models, transfer to multi-table and general tool-use tasks, and ablations showing that simple data and binary rewards outperform more elaborate setups, since the real bottleneck is disciplined tool use rather than reasoning depth.

Oct 09, 2026

Why Your Company Needs a Context Graph (and How to Build It) — Gil Feig, Merge

Gil Feig, co-founder and CTO of Merge, explains why agents need a company-wide context graph rather than a few MCP connections. Using the question of which customers were upset last week, he shows why live API lookups fail across large datasets and why synced local copies, routing, summarization, skills, memory, and traceable provenance are needed to build a reliable company brain.

Oct 09, 2026

Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i

Gabriel Martinez of G2i argues that slop is not AI-written code but work that looks finished before the thinking behind it is done. He connects it to missing judgment, code as a liability, misaligned incentives that reward merging speed over review, oversized PRs and documents that push understanding onto reviewers, and the case for conventions, diagrams as review surfaces, and small changes. He closes by explaining why G2i built ORC to keep humans in control of agent-generated work.

Oct 09, 2026

How MasterClass Builds Multi-Agent AI That Learners Can Trust

MasterClass engineers Mandar and Priya explain how they built NCX, a multi-agent AI tutoring system, and the CoreWeave-based tracing, MCP server, and AgentOps loop that keeps it reliable and self-improving in production.