# Ep 87: Gemini Co-Lead on World Models, RL's Next Domains & Continual Learning

Page: https://stenobird.com/podcast/unsupervised-learning-with-jacob-effron-6041643/ep-87-gemini-co-lead-on-world-models-rl-s-next-domains-continual-learning
Text version: https://stenobird.com/podcast/unsupervised-learning-with-jacob-effron-6041643/ep-87-gemini-co-lead-on-world-models-rl-s-next-domains-continual-learning.md
Podcast: [Unsupervised Learning with Jacob Effron](https://stenobird.com/podcast/unsupervised-learning-with-jacob-effron-6041643)
Published: 2026-05-22T12:50:04+00:00
Episode link: https://unsupervised-learning.simplecast.com/episodes/ep-87-gemini-co-lead-on-world-models-rls-next-domains-continual-learning-GVbiENRb
Audio file: https://cdn.simplecast.com/media/audio/transcoded/b3414ac6-61c8-4752-8722-491e1457c3bf/2c08ad29-5b79-42c0-a40a-6c1af4327f2f/episodes/audio/group/81355dc0-5bcb-445f-b3ab-4bcb1731e2a9/group-item/135fe868-55a2-4836-8b58-6230e30f607b/128_default_tc.mp3?aid=rss_feed&feed=dOSE_bdP
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/unsupervised-learning-with-jacob-effron-6041643/episodes/ep-87-gemini-co-lead-on-world-models-rl-s-next-domains-continual-learning
Duration seconds: 3581

## Resource

Oriol Vinyals, VP of Research at Google DeepMind and co-lead of the Gemini program, joins Jacob the day after Google I/O to unpack the research underpinning Google's latest announcements and where frontier AI is heading. The conversation moves from world models (why Google has uniquely bet on them as a path to AGI, what the "GPT moment" for video and images would look like, and how they connect to robotics and simulation) to agents (the Spark release, why the system and model need to be optimized jointly, and why scaffolding will eventually be written by models themselves). Oriol gets into the mechanics of memory in models, drawing on his cognitive neuroscience background to argue that file-system-style non-parametric memory is more practical than baking memory into weights at serving scale. He shares his views on the limits of RL today (LLMs are data-limited in a way that game-playing RL never was), why training on narrow domains like math and code generalizes surprisingly well, and what a true "Move 37" moment for science or ML research would look like. Throughout, he reflects on the unique advantages of being inside Google (TPU co-design, end-to-end revenue stability, the merger of Brain and DeepMind), the trade-offs between focus and exploration in research orgs, and why he believes AGI in some meaningful sense may already be here, even if the goalposts keep moving.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/unsupervised-learning-with-jacob-effron-6041643/episodes/ep-87-gemini-co-lead-on-world-models-rl-s-next-domains-continual-learning/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/unsupervised-learning-with-jacob-effron-6041643/ep-87-gemini-co-lead-on-world-models-rl-s-next-domains-continual-learning.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.