Best LessWrong Podcasts (2025)

1
“Sonnet 4.5’s eval gaming seriously undermines alignment evals, and this seems caused by training on alignment evals” by Alexa Pan, ryan_greenblatt 35:57

Play Pause

2h ago35:57

35:57

According to the Sonnet 4.5 system card, Sonnet 4.5 is much more likely than Sonnet 4 to mention in its chain-of-thought that it thinks it is being evaluated; this seems to meaningfully cause it to appear to behave better in alignment evaluations. So, Sonnet 4.5's behavioral improvements in these evaluations may partly be driven by growing tendency…

1
Bits and bricks: Oliver Habryka on LessWrong, LightHaven, and community infrastructure 1:14:44

28d ago1:14:44

1:14:44

Patrick McKenzie (patio11) is joined by Oliver Habryka, who runs Lightcone Infrastructure—the organization behind both the LessWrong forum and the Lighthaven conference venue in Berkeley. They explore how LessWrong became one of the most intellectually consequential forums on the internet, the surprising challenges of running a hotel with fractal g…

1
“Publishing academic papers on transformative AI is a nightmare” by Jakub Growiec 7:23

23m ago7:23

7:23

I am a professor of economics. Throughout my career, I was mostly working on economic growth theory, and this eventually brought me to the topic of transformative AI / AGI / superintelligence. Nowadays my work focuses mostly on the promises and threats of this emerging disruptive technology. Recently, jointly with Klaus Prettner, we’ve written a pa…

1
How deposit insurance actually works 33:11

5h ago33:11

33:11

Patrick McKenzie (patio11) reads his Bits about Money essay on deposit insurance, explaining this critical financial infrastructure, with some thoughts on its performance during 2023. He covers what deposit insurance actually covers (and critically, what it doesn't), how fintech users often misunderstand their exposure to counterparty risk, and the…

1
“The Unreasonable Effectiveness of Fiction” by Raelifin 15:03

2h ago15:03

15:03

[Meta: This is Max Harms. I wrote a novel about China and AGI, which comes out today. This essay from my fiction newsletter has been slightly modified for LessWrong.] In the summer of 1983, Ronald Reagan sat down to watch the film War Games, starring Matthew Broderick as a teen hacker. In the movie, Broderick's character accidentally gains access t…

1
“Legible vs. Illegible AI Safety Problems” by Wei Dai 3:29

1h ago3:29

3:29

Some AI safety problems are legible (obvious or understandable) to company leaders and government policymakers, implying they are unlikely to deploy or allow deployment of an AI while those problems remain open (i.e., appear unsolved according to the information they have access to). But some problems are illegible (obscure or hard to understand, o…

1
“Lack of Social Grace is a Lack of Skill” by Screwtape 11:08

7h ago11:08

11:08

1. I have claimed that one of the fundamental questions of rationality is “what am I about to do and what will happen next?” One of the domains I ask this question the most is in social situations. There are a great many skills in the world. If I had the time and resources to do so, I’d want to master all of them. Wilderness survival, automotive re…

1
[Linkpost] “I ate bear fat with honey and salt flakes, to prove a point” by aggliu 1:07

3h ago1:07

1:07

This is a link post. Eliezer Yudkowsky did not exactly suggest that you should eat bear fat covered with honey and sprinkled with salt flakes. What he actually said was that an alien, looking from the outside at evolution, would predict that you would want to eat bear fat covered with honey and sprinkled with salt flakes. Still, I decided to buy a …

1
“What’s up with Anthropic predicting AGI by early 2027?” by ryan_greenblatt 39:25

3h ago39:25

39:25

As far as I'm aware, Anthropic is the only AI company with official AGI timelines[1]: they expect AGI by early 2027. In their recommendations (from March 2025) to the OSTP for the AI action plan they say: As our CEO Dario Amodei writes in 'Machines of Loving Grace', we expect powerful AI systems will emerge in late 2026 or early 2027. Powerful AI s…

1
[Linkpost] “Emergent Introspective Awareness in Large Language Models” by Drake Thomas 3:00

16h ago3:00

3:00

This is a link post. New Anthropic research (tweet, blog post, paper): We investigate whether large language models can introspect on their internal states. It is difficult to answer this question through conversation alone, as genuine introspection cannot be distinguished from confabulations. Here, we address this challenge by injecting representa…

1
[Linkpost] “You’re always stressed, your mind is always busy, you never have enough time” by mingyuan 4:17

1d ago4:17

4:17

This is a link post. You have things you want to do, but there's just never time. Maybe you want to find someone to have kids with, or maybe you want to spend more or higher-quality time with the family you already have. Maybe it's a work project. Maybe you have a musical instrument or some sports equipment gathering dust in a closet, or there's so…

1
“LLM-generated text is not testimony” by TsviBT 19:40

2d ago19:40

19:40

Crosspost from my blog. Synopsis When we share words with each other, we don't only care about the words themselves. We care also—even primarily—about the mental elements of the human mind/agency that produced the words. What we want to engage with is those mental elements. As of 2025, LLM text does not have those elements behind it. Therefore LLM …

1
“Post title: Why I Transitioned: A Case Study” by Fiora Sunshine 17:21

2d ago17:21

17:21

An Overture Famously, trans people tend not to have great introspective clarity into their own motivations for transition. Intuitively, they tend to be quite aware of what they do and don't like about inhabiting their chosen bodies and gender roles. But when it comes to explaining the origins and intensity of those preferences, they almost universa…

1
“The Memetics of AI Successionism” by Jan_Kulveit 21:27

4d ago21:27

21:27

TL;DR: AI progress and the recognition of associated risks are painful to think about. This cognitive dissonance acts as fertile ground in the memetic landscape, a high-energy state that will be exploited by novel ideologies. We can anticipate cultural evolution will find viable successionist ideologies: memeplexes that resolve this tension by fram…

1
“How Well Does RL Scale?” by Toby_Ord 16:11

5d ago16:11

16:11

This is the latest in a series of essays on AI Scaling. You can find the others on my site. Summary: RL-training for LLMs scales surprisingly poorly. Most of its gains are from allowing LLMs to productively use longer chains of thought, allowing them to think longer about a problem. There is some improvement for a fixed length of answer, but not en…

1
“An Opinionated Guide to Privacy Despite Authoritarianism” by TurnTrout 7:59

5d ago7:59

7:59

I've created a highly specific and actionable privacy guide, sorted by importance and venturing several layers deep into the privacy iceberg. I start with the basics (password manager) but also cover the obscure (dodging the millions of Bluetooth tracking beacons which extend from stores to traffic lights; anti-stingray settings; flashing GrapheneO…

1
Home improvement lending with fewer bankers and more computers 28:48

7d ago28:48

28:48

In this episode, Patrick McKenzie reads his essay about the financial infrastructure that makes buying windows painless. When a window installer can originate, underwrite, and fund a $25,000 loan in 15 minutes before leaving your house, it's because four parties—window companies, facilitating platforms, specialized banks, and capital providers—have…

1
“Cancer has a surprising amount of detail” by Abhishaike Mahajan 23:54

6d ago23:54

23:54

There is a very famous essay titled ‘Reality has a surprising amount of detail’. The thesis of the article is that reality is filled, just filled, with an incomprehensible amount of materially important information, far more than most people would naively expect. Some of this detail is inherent in the physical structure of the universe, and the res…

1
249 – Red Heart and IABIED with Max Harms 2:08:06

6d ago2:08:06

2:08:06

We talk with Max Harms on the air for the first time since 2017! He’s got a new book coming out (pre-order your copy here or at Amazon) and we spend about the first half talking about If Anyone Builds It, Everyone Dies. LINKS Max’s first book, Crystal Society Eneasz’s audiobook of about the first two thirds of the first book And the official audiob…

1
“AIs should also refuse to work on capabilities research” by Davidmanheim 6:34

7d ago6:34

6:34

There's a strong argument that humans should stop trying to build more capable AI systems, or at least slow down progress. The risks are plausibly large but unclear, and we’d prefer not to die. But the roadmaps of the companies pursuing these systems envision increasingly agentic AI systems taking over the key tasks of researching and building supe…

1
“On Fleshling Safety: A Debate by Klurl and Trapaucius.” by Eliezer Yudkowsky 2:22:21

8d ago2:22:21

2:22:21

(23K words; best considered as nonfiction with a fictional-dialogue frame, not a proper short story.) Prologue: Klurl and Trapaucius were members of the machine race. And no ordinary citizens they, but Constructors: licensed, bonded, and insured; proven, experienced, and reputed. Together Klurl and Trapaucius had collaborated on such famed artifice…

1
“EU explained in 10 minutes” by Martin Sustrik 16:47

12d ago16:47

16:47

If you want to understand a country, you should pick a similar country that you are already familiar with, research the differences between the two and there you go, you are now an expert. But this approach doesn’t quite work for the European Union. You might start, for instance, by comparing it to the United States, assuming that EU member countri…

1
“Cheap Labour Everywhere” by Morpheus 3:38

12d ago3:38

3:38

I recently visited my girlfriend's parents in India. Here is what that experience taught me: Yudkowsky has this facebook post where he makes some inferences about the economy after noticing two taxis stayed in the same place while he got his groceries. I had a few similar experiences while I was in India, though sadly I don't remember them in enoug…

1
[Linkpost] “Consider donating to AI safety champion Scott Wiener” by Eric Neyman 2:35

12d ago2:35

2:35

This is a link post. Written in my personal capacity. Thanks to many people for conversations and comments. Written in less than 24 hours; sorry for any sloppiness. It's an uncanny, weird coincidence that the two biggest legislative champions for AI safety in the entire country announced their bids for Congress just two days apart. But here we are.…

1
“Which side of the AI safety community are you in?” by Max Tegmark 4:18

13d ago4:18

4:18

In recent years, I’ve found that people who self-identify as members of the AI safety community have increasingly split into two camps: Camp A) "Race to superintelligence safely”: People in this group typically argue that "superintelligence is inevitable because of X”, and it's therefore better that their in-group (their company or country) build i…

1
“Doomers were right” by Algon 4:35

12d ago4:35

4:35

There's an argument I sometimes hear against existential risks, or any other putative change that some are worried about, that goes something like this: 'We've seen time after time that some people will be afraid of any change. They'll say things like "TV will destroy people's ability to read", "coffee shops will destroy the social order","machines…

1
Talking to the Bank of England about systemic risk and systems engineering 1:31:37

14d ago1:31:37

1:31:37

Patrick McKenzie (@patio11) shares his remarks to the Bank of England on critical vulnerabilities in financial infrastructure. Drawing from the July 2024 CrowdStrike outage which brought down teller systems at major US banks, Patrick discusses how regulatory guidance inadvertently created dangerous software monocultures. He also examines the stable…

1
“Do One New Thing A Day To Solve Your Problems” by Algon 3:21

13d ago3:21

3:21

People don't explore enough. They rely on cached thoughts and actions to get through their day. Unfortunately, this doesn't lead to them making progress on their problems. The solution is simple. Just do one new thing a day to solve one of your problems. Intellectually, I've always known that annoying, persistent problems often require just 5 secon…

LessWrong Podcasts

1
LessWrong (Curated & Popular)

LessWrong

1
The Bayesian Conspiracy

The Bayesian Conspiracy

1
Complex Systems with Patrick McKenzie (patio11)

Patrick McKenzie

1
Heart of the Matter

Bryan Davis & Jay Kannaiyan

1
“Sonnet 4.5’s eval gaming seriously undermines alignment evals, and this seems caused by training on alignment evals” by Alexa Pan, ryan_greenblatt 35:57

1
Bits and bricks: Oliver Habryka on LessWrong, LightHaven, and community infrastructure 1:14:44

1
“Publishing academic papers on transformative AI is a nightmare” by Jakub Growiec 7:23

1
How deposit insurance actually works 33:11

1
“The Unreasonable Effectiveness of Fiction” by Raelifin 15:03

1
“Legible vs. Illegible AI Safety Problems” by Wei Dai 3:29

1
“Lack of Social Grace is a Lack of Skill” by Screwtape 11:08

1
[Linkpost] “I ate bear fat with honey and salt flakes, to prove a point” by aggliu 1:07

1
“What’s up with Anthropic predicting AGI by early 2027?” by ryan_greenblatt 39:25

1
[Linkpost] “Emergent Introspective Awareness in Large Language Models” by Drake Thomas 3:00

1
[Linkpost] “You’re always stressed, your mind is always busy, you never have enough time” by mingyuan 4:17

1
“LLM-generated text is not testimony” by TsviBT 19:40

1
“Post title: Why I Transitioned: A Case Study” by Fiora Sunshine 17:21

1
“The Memetics of AI Successionism” by Jan_Kulveit 21:27

1
“How Well Does RL Scale?” by Toby_Ord 16:11

1
“An Opinionated Guide to Privacy Despite Authoritarianism” by TurnTrout 7:59

1
Home improvement lending with fewer bankers and more computers 28:48

1
“Cancer has a surprising amount of detail” by Abhishaike Mahajan 23:54

1
249 – Red Heart and IABIED with Max Harms 2:08:06

1
“AIs should also refuse to work on capabilities research” by Davidmanheim 6:34

1
“On Fleshling Safety: A Debate by Klurl and Trapaucius.” by Eliezer Yudkowsky 2:22:21

1
“EU explained in 10 minutes” by Martin Sustrik 16:47

1
“Cheap Labour Everywhere” by Morpheus 3:38

1
[Linkpost] “Consider donating to AI safety champion Scott Wiener” by Eric Neyman 2:35

1
“Which side of the AI safety community are you in?” by Max Tegmark 4:18

1
“Doomers were right” by Algon 4:35

1
Talking to the Bank of England about systemic risk and systems engineering 1:31:37

1
“Do One New Thing A Day To Solve Your Problems” by Algon 3:21