Subvolts take Former OpenAI architect details AI scaling limits and the path to AGI, offering candid research insights. Jan 29, 2026 - Developers
- Learn about the fundamental limits of current AI scaling paradigms and the research directions needed for AGI.
- Business
- Understand the potential timeline for significant AI advancements, particularly in robotics, and societal automation impacts.
- Everyday users
- Gain insight into the future of AI development from a researcher who helped build foundational models, including what to expect soon.
A former key architect from OpenAI shares his candid views on AI's future, research challenges, and his departure.
Ep 81: Ex-OpenAI Researcher On Why He Left, His Honest AGI Timeline, & The Limits of Scaling RL
Audio streams from by Redpoint Ventures. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
This episode of Unsupervised Learning features former OpenAI researcher Jerry Tworek discussing AI scaling limits, the necessity of continual learning for AGI, and his departure from OpenAI. He shares insights on research convergence, competitive dynamics, and predicts near-term robotics advancements while warning about societal impacts of automation.
A former top AI builder explains why current AI methods hit limits and what new abilities are needed for true artificial intelligence. He shares why he left OpenAI and what to expect soon in AI, especially for robots.
This episode offers a candid look at the challenges and future of AI development from a researcher who helped build key OpenAI models. It's for anyone interested in the practical limitations of current AI paradigms and the path toward AGI. Tworek's perspective on scaling and continual learning directly relates to ongoing research at major labs like OpenAI, Google Gemini, and Anthropic, who are all pushing the boundaries of model capabilities.
- HostJacob Effron
- HostJordan Segall
- GuestJerry Tworekkey architect, OpenAI
- Guest: Jerry Tworek
- Show: Unsupervised Learning
- Episode Title: Ep 81: Ex-OpenAI Researcher On Why He Left, His Honest AGI Timeline, & The Limits of Scaling RL
- Length: 63 minutes
- Guest's Contributions: Key architect behind OpenAI's o1, o3, and Codex models
- Robotics Prediction: 2-3 years
This episode features Jerry Tworek, a key architect behind OpenAI's breakthrough reasoning models (o1, o3) and Codex, discussing the current state and future of AI. Jerry explores the real limits and promise of scaling pre-training and reinforcement learning, arguing that while these paradigms deliver predictable improvements, they're fundamentally constrained by data availability and struggle with generalization beyond their training objectives. He reveals his updated belief that continual learning—the ability for models to update themselves based on failure and work through problems autonomously—is necessary for AGI, as current models hit walls and become "hopeless" when stuck. Jerry discusses the convergence of major labs toward similar approaches driven by economic forces, the tension between exploration and exploitation in research, and why he left OpenAI to pursue new research directions. He offers candid insights on the competitive dynamics between labs, the focus required to win in specific domains like coding, what makes great AI researchers, and his surprisingly near-term predictions for robotics (2-3 years) while warning about the societal implications of widespread work automation that we're not adequately preparing for.
(0:00) Intro
(1:26) Scaling Paradigms in AI
(3:36) Challenges in Reinforcement Learning
(11:48) AGI Timelines
(18:36) Converging Labs
(25:05) Jerry’s Departure from OpenAI
(31:18) Pivotal Decisions in OpenAI’s Journey
(35:06) Balancing Research and Product Development
(38:42) The Future of AI Coding
(41:33) Specialization vs. Generalization in AI
(48:47) Hiring and Building Research Teams
(55:21) Quickfire
With your co-hosts:
@jacobeffron
- Partner at Redpoint, Former PM Flatiron Health
@patrickachase
- Partner at Redpoint, Former ML Engineer LinkedIn
@ericabrescia
- Former COO Github, Founder Bitnami (acq’d by VMWare)
@jordan_segall
- Partner at Redpoint
Jacob is an AI investor at Redpoint Ventures. He's led Redpoint's investments in companies like Abridge, Physical Intelligence & Legora. Follow Jacob on Twitter (@jacobeffron).
On Unsupervised Learning we probe the sharpest minds in AI in search for the truth about what's real today, what will be real in the future and what it all means for businesses and the world. If you're a builder, researcher or investor navigating the AI world, this podcast will help you deconstruct and understand the most important breakthroughs and see a clearer picture of reality. Subscribe to this show to stay up to date on our latest episodes.
- AGI
- Jerry Tworek
- Artificial General Intelligence, or AI that can perform any intellectual task a human can.
- scaling pre-training
- Jerry Tworek
- Making AI models better by training them on more data and with more computing power.
- reinforcement learning
- Jerry Tworek
- An AI training method where models learn by trial and error, receiving rewards for good actions.
- continual learning
- Jerry Tworek
- The ability for AI models to learn and update themselves over time based on new experiences and failures.
Same guests, other shows
49 minBuilding the Automated AGI Lab: Core Automation's Jerry Tworek and Rohan Anil
WithJerry Tworek, Rohan Anil
1 h 16How GPT-5 Thinks — OpenAI VP of Research Jerry Tworek
WithJerry Tworek
Questions people ask
- Why did Jerry Tworek leave OpenAI?
- He left to pursue new research directions outside of OpenAI's current focus.
- What is Jerry Tworek's updated belief for achieving AGI?
- He believes continual learning, where models update themselves autonomously, is necessary for AGI.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to by Redpoint Ventures.