reinforcement learning, TLexDR

The neighbourhood: reinforcement learning and the ideas it travels with. Drag to roam, click a star for the episode, click a neighbour to travel.

Drag to roam · scroll to zoom · click a neighbour to travel · click a star for the episode

From foundational to frontier

Climb the spectrum. The most accessible conversations come first.

Start here

ACCESSIBLECOREFRONTIER

Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics

1h 1m

05-28-26

Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics

Coming soon

Michael Littman: Reinforcement Learning and the Future of AI

1h 56m

05-28-26

Michael Littman: Reinforcement Learning and the Future of AI

Coming soon

Juergen Schmidhuber: Godel Machines, Meta-Learning, and LSTMs

1h 19m

05-28-26

Juergen Schmidhuber: Godel Machines, Meta-Learning, and LSTMs

Coming soon

David Silver: AlphaGo, AlphaZero, and Deep Reinforcement Learning

1h 48m

05-28-26

David Silver: AlphaGo, AlphaZero, and Deep Reinforcement Learning

Coming soon

Sergey Levine: Robotics and Machine Learning

1h 37m

05-28-26

Sergey Levine: Robotics and Machine Learning

Coming soon

Matt Botvinick: Neuroscience, Psychology, and AI at DeepMind

2h

05-28-26

Matt Botvinick: Neuroscience, Psychology, and AI at DeepMind

Coming soon

42m

05-28-26

Yoshua Bengio: Deep Learning

Coming soon

1h 25m

05-28-26

Greg Brockman: OpenAI and AGI

Coming soon

Pieter Abbeel: Deep Reinforcement Learning

42m

05-28-26

Pieter Abbeel: Deep Reinforcement Learning

Coming soon

The lexicon

Every term the guests lean on, in plain language. Read one in full, or filter to find it.

12

Bitter Lesson

Rich Sutton's argument that simple algorithms leveraging computation have driven the most significant AI advancements.

capped-profit model

A business structure that limits investor returns to align with nonprofit missions.

credit_assignment

The process of determining which components of a neural network are responsible for specific outputs.

disentangled_representations

AI's ability to separate and understand individual variables within data.

Hierarchical Planning

A method in AI for breaking down complex tasks into smaller, manageable sub-tasks.

LSTMs

Long Short-Term Memory networks, a type of recurrent neural network used for tasks requiring memory of past events.

Markov Decision Processes (MDPs)

Mathematical frameworks for modeling decision-making in situations where outcomes are partly random.

meta learning

A process where one learning algorithm gives rise to another, often spontaneously in neural networks.

meta-learning

A process where machines recursively improve their own learning algorithms.

Monte Carlo tree search

An algorithm used to make decisions in game theory, involving random sampling to determine the best move.

Moravec's paradox

Tasks easy for humans are hard for machines and vice versa, highlighting AI development challenges.

off-policy reinforcement learning

Learning from actions taken by different policies or systems, not previously observed in the dataset.

+5 more terms unlock the full lexicon with Read.

What the corpus says

The throughline across every conversation that touches this idea.

AlphaGo's victory in Go marked a significant advancement in AI, showcasing the power of reinforcement learning and self-play.

Michael Littman · Michael Littman: Reinforcement Learning and the Future of AI

Reinforcement learning systems struggle with human interaction due to high costs and low bandwidth, limiting their development.

Michael Littman · Michael Littman: Reinforcement Learning and the Future of AI

Rich Sutton's 'Bitter Lesson' highlights that simple algorithms leveraging computation have driven major AI advancements.

Michael Littman · Michael Littman: Reinforcement Learning and the Future of AI

Self-driving cars face challenges in understanding social cues, which are crucial for safe driving.

Michael Littman · Michael Littman: Reinforcement Learning and the Future of AI

The exponential growth of technology may reach a limit, leading to diminishing returns rather than endless improvement.

Michael Littman · Michael Littman: Reinforcement Learning and the Future of AI

Robots excel in controlled environments but struggle in unpredictable ones due to a lack of common sense and adaptability.

Sergey Levine · Sergey Levine: Robotics and Machine Learning

Reinforcement learning is evolving from utility maximization to exploration-first approaches, crucial for robotics development.

Sergey Levine · Sergey Levine: Robotics and Machine Learning

Simulation is vital for reinforcement learning but can limit progress if not complemented by real-world data.

Sergey Levine · Sergey Levine: Robotics and Machine Learning

Sergey Levine argues that nefarious humans are a bigger existential threat than AI systems themselves.

Sergey Levine · Sergey Levine: Robotics and Machine Learning

Combining perception and control in robotics can outperform traditional modular approaches, as seen in end-to-end reinforcement learning.

Sergey Levine · Sergey Levine: Robotics and Machine Learning

Meta learning in AI can emerge spontaneously in recurrent neural networks, creating new learning algorithms from network dynamics.

Matt Botvinick · Matt Botvinick: Neuroscience, Psychology, and AI at DeepMind

Dopamine's role in reinforcement learning mirrors temporal difference learning, suggesting a neural basis for AI techniques.

Matt Botvinick · Matt Botvinick: Neuroscience, Psychology, and AI at DeepMind

Voices on reinforcement learning

12 standout quotes from across the corpus.

Go read

39 books and papers cited across these episodes.

For the specialist

What experts find new

19 expert-level takeaways for a specialist reader.

At the frontier

Still unresolved

13 open questions flagged across these conversations.

The thinkers

Who takes this idea on, by how often they return to it.

1 DS

1 GB

1 LK

1 MB

1

1 SL

1 YB

1 JS

1

Adjacent ideas

robotics3 self-play3 AGI2 AI safety2 abstraction1 AI alignment1 AI breakthroughs1 AI creativity1 AI expansion1 AI reasoning1 credit assignment1 deep learning1