The a16z Show · 25 September 2025
At the source: Andreessen Horowitz ↗ · audio
What comes after vibe coding? Maybe vibe researching. OpenAI’s Chief Scientist, Jakub Pachocki, and Chief Research Officer, Mark Chen, join a16z general partners Anjney Midha and Sarah Wang to go deep on GPT-5—how they fused fast replies with long-horizon reasoning, how they measure progress once benchmarks saturate, and why reinforcement learning keeps surprising skeptics. They explore agentic systems (and their stability tradeoffs), coding models that change how software gets made, and the bigger bet: an automated researcher that can generate new ideas with real economic impact. Plus: how they prioritize compute, hire “cave-dweller” talent, protect fundamental research inside a product company, and keep pace without chasing every shiny demo. Timecodes: 0:00 Introduction & Goals of Automated Researcher 0:43 The Evolution of Reasoning in AI 1:46 Evaluations: From Benchmarks to Real-World Impact 5:15 Surprising Capabilities of GPT-5 6:56 The Research Roadmap: Next 1, 2, 5 Years 7:46 Long-Horizon Agency & Model Memory 9:44 Reasoning in Open-Ended Domains 11:18 The Role and Progress of Reinforcement Learning 13:14 Reward Modeling & Best Practices 14:21 The New Codex: Real-World Coding 16:20 AI vs. Human Coding: The New Default 20:07 What Makes a Great Researcher?
As published with the episode at The a16z Show ↗
No summary of this episode yet.