Trending Episodes

The Andrej Karpathy episode.

During this interview, Andrej explains why reinforcement learning is terrible (but everything else is much worse), why AGI will just blend into the previous ~2.5 centuries of 2% GDP growth, why self driving took so long to crack, and what he sees as the future of education.

It was a pleasure chatting with him.

Watch on YouTube; read the transcript.

Sponsors

* Labelbox helps you get data that is more detailed, more accurate, and higher signal than you could get by default, no matter your domain or training paradigm. Reach out today at labelbox.com/dwarkesh

* Mercury helps you run your business better. It’s the banking platform we use for the podcast — we love that we can see our accounts, cash flows, AR, and AP all in one place. Apply online in minutes at mercury.com

* Google’s Veo 3.1 update is a notable improvement to an already great model. Veo 3.1’s generations are more coherent and the audio is even higher-quality. If you have a Google AI Pro or Ultra plan, you can try it in Gemini today by visiting https://gemini.google

Timestamps

(00:00:00) – AGI is still a decade away

(00:29:45) – LLM cognitive deficits

(00:40:05) – RL is terrible

(00:49:38) – How do humans learn?

(01:06:25) – AGI will blend into 2% GDP growth

(01:17:36) – ASI

(01:32:50) – Evolution of intelligence & culture

(01:42:55) - Why self driving took so long

(01:56:20) - Future of education

After getting the red-carpet treatment at the UN in New York last week, the former al-Qaeda fighter who now leads Syria is about to hold an election. But is Syria’s new president, Ahmed al-Sharaa, really about to transition the country into democracy? Or does he have other plans? The BBC's senior international correspondent Orla Guerin joins us from Damascus, where she’s been speaking to Syrians about the country’s future.

With Asma Khalid in DC, Tristan Redman in London, and the backing of the BBC’s international newsroom, The Global Story brings clarity to politics, business and foreign policy in a time of connection and disruption. For more episodes, just search 'The Global Story' wherever you get your BBC Podcasts. . Producers: Cat Farnsworth and Valerio Esposito Executive producer: James Shield Mix: Travis Evans Senior News Editor: China Collins

Richard Sutton is the father of reinforcement learning, winner of the 2024 Turing Award, and author of The Bitter Lesson. And he thinks LLMs are a dead end.

After interviewing him, my steel man of Richard’s position is this: LLMs aren’t capable of learning on-the-job, so no matter how much we scale, we’ll need some new architecture to enable continual learning.

And once we have it, we won’t need a special training phase — the agent will just learn on-the-fly, like all humans, and indeed, like all animals.

This new paradigm will render our current approach with LLMs obsolete.

In our interview, I did my best to represent the view that LLMs might function as the foundation on which experiential learning can happen… Some sparks flew.

A big thanks to the Alberta Machine Intelligence Institute for inviting me up to Edmonton and for letting me use their studio and equipment.

Enjoy!

Watch on YouTube; listen on Apple Podcasts or Spotify.

Sponsors

* Labelbox makes it possible to train AI agents in hyperrealistic RL environments. With an experienced team of applied researchers and a massive network of subject-matter experts, Labelbox ensures your training reflects important, real-world nuance. Turn your demo projects into working systems at labelbox.com/dwarkesh

* Gemini Deep Research is designed for thorough exploration of hard topics. For this episode, it helped me trace reinforcement learning from early policy gradients up to current-day methods, combining clear explanations with curated examples. Try it out yourself at gemini.google.com

* Hudson River Trading doesn’t silo their teams. Instead, HRT researchers openly trade ideas and share strategy code in a mono-repo. This means you’re able to learn at incredible speed and your contributions have impact across the entire firm. Find open roles at hudsonrivertrading.com/dwarkesh

Timestamps

(00:00:00) – Are LLMs a dead end?

(00:13:04) – Do humans do imitation learning?

(00:23:10) – The Era of Experience

(00:33:39) – Current architectures generalize poorly out of distribution

(00:41:29) – Surprises in the AI field

(00:46:41) – Will The Bitter Lesson still apply post AGI?

(00:53:48) – Succession to AIs

From Apple News In Conversation: The United States has the highest maternal mortality rate among high-income nations — and, according to the CDC, more than 80% of maternal deaths are preventable. In her new book, Unbearable: Five Women and the Perils of Pregnancy in America, journalist Irin Carmon follows families as they navigate fertility struggles, pregnancy, birth, and loss within a health-care system that too often fails them. Carmon sat down with Apple News In Conversation host Shumita Basu to talk about how the history of maternal health care in the U.S. continues to shape the lives of pregnant people today.

Pre-order the Planet Money book here for your free gift. 

Our sister show, The Indicator, is chronicling the evolving business of crime for its Vice Week series. Today, we bring to you two cases of crime in the age of AI. 

First, cybercriminals are using our own voices against us. Audio deepfake scams are picking up against individuals but also against businesses. We hear from a bank on how they’re adapting defenses, and find out how the new defenses are a game of AI vs AI. 

Then, we move over to the stock market to witness AI market manipulation. A new breed of trading bots behave differently. They could collude with each other, even without human involvement or instruction, so researchers are asking how to think about blame, and regulation in a world of more sophisticated trading bots. That’s assuming regulators could even keep up with the tech in the first place. 

Indicator Vice Series 
Head to The Indicator from Planet Money podcast feed for the latest on the Indicator Vice Series including an episode on data breaches . If you don’t already subscribe, check it out. Each episode explains one slice of the economy connected to the news recently, always in 10 minutes or less. 

Subscribe to Planet Money+ 

Listen free: Apple Podcasts, Spotify, the NPR app or anywhere you get podcasts.

Facebook / Instagram / TikTok / Our weekly Newsletter. 

This episode is hosted by Darian Woods, Adrian Ma, and Wailin Wong. These episodes of The Indicator were originally produced by Cooper Katz McKim and engineered by Robert Rodriguez. They were fact-checked by Sierra Juarez. Kate Concannon is The Indicator’s editor. Alex Goldmark is the Executive Producer. 

Music: NPR Source Audio - “Diamond High” 

Learn more about sponsor message choices: podcastchoices.com/adchoices

NPR Privacy Policy