TalkRL: The Reinforcement Learning Podcast

About TalkRL: The Reinforcement Learning Podcast

TalkRL podcast is All Reinforcement Learning, All the Time. In-depth interviews with brilliant people at the forefront of RL research and practice. Guests from places like MILA, OpenAI, MIT, DeepMind, Berkeley, Amii, Oxford, Google Research, Brown, Waymo, Caltech, and Vector Institute. Hosted by Robin Ranjit Singh Chauhan.

Publisher

Robin Ranjit Singh Chauhan

Category

technology · technology

Language

en

Explicit

No

First episode

1 Aug 2019

Latest episode

14 Aug 2026

Latest episodes

24 of 76 episodes have a transcript.

1. 

   14 Aug 2026TRANSCRIPT

   Thomas Frost on Clinical RL with Natural Timings

   Dr. Thomas Frost, an emergency physician and PhD candidate, discusses applying reinforcement learning with natural timings to healthcare settings in this interview.

2. 

   3 Jan 2026TRANSCRIPT

   Joseph Modayil of Openmind Research Institute @ RLC 2025

   Joseph Modayil is the Founder, President & Research Director of Openmind Research Institute. Featured References Openmind Research Institute (https://www.openmindresearch.org/) The Alberta Plan for AI Research (https://arxiv.org/abs/2208.11173) Richard S. Sutton, Michael Bowling, Patrick M. Pilarski Additional References Joseph Modayil on Google Scholar (https://scholar.google.co.in/citations?user=G3pvUNEAAAAJ&hl=ja) Joseph Modayil Homepage (https://josephmodayil.com/)

3. 

   10 Nov 2025TRANSCRIPT

   Danijar Hafner on Dreamer v4

   Danijar Hafner was a Research Scientist at Google DeepMind until recently. Featured References Training Agents Inside of Scalable World Models (https://arxiv.org/abs/2509.24527) [ blog (https://danijar.com/project/dreamer4/) ] Danijar Hafner, Wilson Yan, Timothy Lillicrap One Step Diffusion via Shortcut Models (https://arxiv.org/abs/2410.12557) Kevin Frans, Danijar Hafner, Sergey Levine, Pieter Abbeel Action and Perception as Divergence Minimization (https://arxiv.org/abs/2009.01791) [ blog (https://danijar.com/project/apd/) ] Danijar Hafner, Pedro A. Ortega, Jimmy Ba, Thomas Parr, Karl Friston, Nicolas Heess Additional References Mastering Diverse Domains through World Models (https://arxiv.org/abs/2301.04104v1) [ blog (https://danijar.com/project/dreamerv3/) ] DreaverV3l Danijar Hafner, Jurgis Pasukonis, Jimmy Ba, Timothy Lillicrap Mastering Atari with Discrete World Models (https://arxiv.org/abs/2010.02193) [ blog (https://danijar.com/project/dreamerv2/) ] DreaverV2 ; Danijar Hafner, Timothy Lillicrap, Mohammad Norouzi, Jimmy Ba Dream to Control: Learning Behaviors by Latent Imagination (https://arxiv.org/abs/1912.01603) [ blog (https://danijar.com/project/dreamer/) ] Dreamer ; Danijar Hafner, Timothy Lillicrap, Jimmy Ba, Mohammad Norouzi Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos (https://arxiv.org/abs/2206.11795) [ Blog Post (https://openai.com/research/vpt) ], Baker et al

Inside a recent episode

Thomas Frost on Clinical RL with Natural Timings

Published 14 Aug 2026 · Transcript excerpt

[…] So a really simple example of that would be maybe I'm treating patients at random intervals, 30 minutes, 90 minutes. But when I pre-process my data, the model just sees fixed regular decisions every one hour. Well, from the model's point of view, that's the environment it expects to be deployed into is having decisions once an hour, every hour. And so when you start exposing it to an environment that isn't like that, it may well do poorly. But I think the most interesting bit about that is that if you are trying to predict how your model is going to do and you do that analysis on your pre-processed data, then your predictions are also wrong because you're predicting how it will do within this fictional pre-processed data set, not how it will do in your real world, messy, regular decision making. […]

Reach and audience

Public platform figures. Ratings count people who left a rating, not total listeners.

Apple Podcasts (US)

4.9 / 5

30 ratings

Spotify

5.0 / 5

74 ratings

Podcast Authority Score: 62 / 100

A composite of feed quality, social presence, YouTube performance and engagement. Read the methodology.

Quality

86

Social presence

0

YouTube

0

Engagement

92

Contact TalkRL: The Reinforcement Learning Podcast

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.

Shows with a similar audience

• 

  The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

• 

  Machine Learning Street Talk (MLST)

• 

  Training Data

• 

  Super Data Science: ML & AI Podcast with Jon Krohn

• 

  No Priors: Artificial Intelligence | Machine Learning | Technology | Startups

• 

  Google DeepMind: The Podcast

Questions about TalkRL: The Reinforcement Learning Podcast