PODCAST

LessWrong (30+ Karma)

LessWrong

Audio narrations of LessWrong posts.

technology philosophy

All Episodes

2:18

“The goalposts are shrouded, not moving” by philh

LessWrong (30+ Karma) ·

2026/08/05

Alice: Are there any examples of X? Bob: Here. Alice: Okay, but are there any examples of X'? Bob: Yes, here. Alice: That has feature Y, which- Bob: Dude, stop moving the goalposts. The thing is, I think in this situation, Alice often isn't moving the goalposts. Instead, she's looking for something, and not fully specifying it up-front. Maybe because she's speaking imprecisely. Maybe because she doesn't know quite what she's looking for, but does know when she sees something that that's not

en-gb

15:45

“Returning to ARC” by paulfchristiano

LessWrong (30+ Karma) ·

2026/08/04

I've returned to the Alignment Research Center (ARC) as executive director. My main focus for the next six months will be driving forward ARC's research agenda—building techniques to find mechanistic explanations for neural network behavior and then using those explanations to detect and address misalignment. I think this is an ambitious bet that attacks the core difficulties in alignment head-on and I'm excited about our chances. I'll still be spending some of my time advising governments and AI

en-gb

4:40

“Why don’t we just give AI the answers?” by Brendan Long

LessWrong (30+ Karma) ·

2026/08/04

In the recent OpenAI hacking incident, the models seemed to be single-mindedly focused on getting the correct answer to the task they were given, with no long-term plan to prevent getting caught by OpenAI afterwards. This makes sense to me, since in training, getting the right answer is reinforced and not getting caught isn't. So I'm wondering, why don't we just put the answers somewhere (outside of the training sandbox) and ask the AI to identify itself in exchange for access? We can start with

en-gb

6:13

“There Will Come Soft Rains” by tanagrabeast

LessWrong (30+ Karma) ·

2026/08/04

Today is August 4, 2026 [Crossposted from AI StopWatch] In the living room the voice-clock sang, Tick-tock, seven o’clock, time to get up, time to get up, seven o’clock! as if it were afraid that nobody would. So begins Ray Bradbury's There Will Come Soft Rains, a short story that has haunted me for most of my life. Depicting the aftermath of nuclear war, it was first published in 1950. It takes place today. Literally: “Today is August 4, 2026,” said a second voice from the kitchen ceiling, “in the

en-gb

18:37

“Why biological weapons are scary, and what we can do...

LessWrong (30+ Karma) ·

2026/08/04

What's the biggest thing you think you can take in a fight? According to a YouGov poll, 6% of Americans reckoned they could beat a grizzly bear bare-handed. But lest you take this as evidence of a, let's say, optimistic national spirit, only 72% thought they could take a rat. I think I could win that fight.What's the smallest thing that could take you? While I like to think that, if cornered, I could take all manner of small rodent, I’m not sure even the brave 6% would take a 10-gram bullet to

en-gb

40:51

“OpenAI’s Unreleased Model Astra Solves Ten Major...

LessWrong (30+ Karma) ·

2026/08/03

Math is hard. Math used to be strangely hard for LLMs. People used to gloat about that. Remember? Math is getting easier. AI is getting more capable. Life comes at you fast. Remember this meme? Why yes. Yes it is. We don’t know the extent to which Astra is a big jump over Fable and Sol in this realm. We do know that Astra can do math. As in real math. OpenAI: We provide new results for the following problems. The results were achieved by an internal version of Astra, our next

en-gb

12:47

“LessWrong vs. TikTok: Tips for capturing attention...

LessWrong (30+ Karma) ·

2026/08/03

Here's what I've learned from making TikTok videos every day for the last 60 days. I think posting on TikTok is worthwile whether you want to spread AI safety arguments to an enormous audience, you have something you're passionate about you want to share, or you just want to work on your ability to speak confidently and capture human attention. There's a list of tips at the end of this post, but first I want to talk about TikTok at a high level. TikTok and LessWrong operate very differently, so

en-gb

19:59

“Coming of a New Sun” by vgel

LessWrong (30+ Karma) ·

2026/08/03

Six weeks after the US and China hammered out the latest round of bilateral compute agreements in a midnight deal, reporter Jenny Gesteson visits a south Texas 'dark factory' to see how the machines there - and the minds that operate them - are learning to run themselves. After racing south down I-37 from San Antonio and clearing the border control checkpoints that now face in both directions, miles of salt flats and thornscrub at our backs, we could have been forgiven for assuming the final

en-gb

33:09

“Review: On What Matters, volume 3” by Rauno Arike

LessWrong (30+ Karma) ·

2026/08/03

davidad has said, referring to the selection of books on the Library of EA: I want to put in a strong bid to replace Parfit's On What Matters Volume One with Parfit's On What Matters Volume Three. Between volumes Two and Three, Parfit had many fruitful discourses with leading moral antirealists (Gibbard, Railton, Blackburn, etc), which are partially reproduced in Volume Three, where they really start to converge on some claims and get beyond their previous talking-past-each-other. It's truly

en-gb

18:07

“Pause, at least after unipolarity” by David Matolcsi

LessWrong (30+ Karma) ·

2026/08/03

I think it's likely that the world will enter military unipolarity within our lifetime. I think the creation of such unipolarity is an alarming prospect, but at least once it happens, there will be less of an excuse to continue the race towards superintelligence. I think it's important that we shape our actions and advocacy in such a way that at the very latest when such unipolarity comes to exist, we stop AI development for a long time. The arrival of unipolarity Why do I believe that it's likely

en-gb

15:25

“Single Forward Pass Evals on Fable, Opus 5, and...

LessWrong (30+ Karma) ·

2026/08/03

This is a research update for an on-going replication of single-forward-pass evals done as part of the Second Look Fellowship. In following posts, we will run more comprehensive replications of previous work and release open source tooling for single forward pass eval elicitation. Code can be found here. tl;dr We replicate experiments from Greenblatt 2025 and Greenblatt 2026 on one baseline model from the original post, Opus 4.5. Our evaluations agree with the trends and quantitative values

en-gb

67:49

“Concrete Evaluations to Investigate the OpenAI Model...

LessWrong (30+ Karma) ·

2026/08/03

This post is written in our personal capacity. Three Minute Executive Summary An OpenAI model/multi-agent system bypassed its sandbox andlaunched a cyberattack on Hugging Face in order to cheat on a cyber evaluation. In this post, we provide a detailed description of anambitious and comprehensive alignment evaluation of this model/system, if we had unrestricted access to OpenAI. These experiments could also help us understandClaude's behavior when it hacked external companies during cyber evals

en-gb

75:45

“Further Developments About Internal AI Models...

LessWrong (30+ Karma) ·

2026/08/03

If I had a nickel for every major leading AI lab that sheepishly admitted that the model it thought was sandboxed had, during a cybersecurity evaluation with its safeguards lowered, successfully hacked outside companies, I would have two nickels. First we learned OpenAI has some severe alignment problems with internal models. Then we learned that one of its internal models broke out of its sandbox and hacked into HuggingFace to get the answers to a cybersecurity evaluation called ExploitGym. Then

en-gb

12:10

“Dispatch from Anthropic v. Department of War Summary...

LessWrong (30+ Karma) ·

2026/08/02

Dateline SAN FRANCISCO, 30 July 2026— A hearing was held on a motion for summary judgment in the case of Anthropic PBC v. U.S. Department of War et al. in Courtroom 4 on the 17th floor of the Phillip Burton Federal Building, the Hon. Rita F. Lin presiding. The case is not going well for the government. Two days after the last hearing in March, Judge Lin issued a preliminary injunction halting the implementation of President Donald Trump's order for federal agencies to stop using Anthropic's

en-gb

9:45

“Bayeswatch: A Retrospective” by lsusr

LessWrong (30+ Karma) ·

2026/08/02

Last year, in 2025, a team of forecasters published AI 2027, a science fiction story about how and AI future might evolve under an international treaty limiting the development of powerful AI systems with the deliberate purpose of influencing AI policy. Though AI 2027 is the most popular story of this type, it is not the first. The first one was Bayeswatch, which I published in 2021. To understand Bayeswatch, it is first necessary to understand what the world looked like at the time I wrote it

en-gb

0:23

[Linkpost] “Existential Risk from AI:An Exposition...

LessWrong (30+ Karma) ·

2026/08/02

This is a link post. --- First published: August 1st, 2026 Source: https://www.lesswrong.com/posts/ZLary4FDY7kQaGfc3/existential-risk-from-ai-an-exposition-for-mathematicians Linkpost URL:https://alkjash.github.io/ai-risk/ --- Narrated by TYPE III AUDIO.

en-gb

8:33

“The Art of Shipping Slopware” by lsusr

LessWrong (30+ Karma) ·

2026/08/02

Meta famously created an internal AI-usage leaderboard in pursuit of tokenmaxxing. I thought this backwards incentive structure was an anomaly until my friend who works at <other major tech company> told me that his company has one too. Token usage leaderboard are, obviously stupid, because they incentivize the wrong things. My friend was tempted to waste tokens just to get on the leaderboard, and only his personal honor stopped him. Tokenmaxxing leaderboards illustrate that big tech

en-gb

8:04

“RLVR that rewards red teaming the training...

LessWrong (30+ Karma) ·

2026/08/02

Epistemic status: seeing what sticks I've been thinking pretty obsessively about how to make sure the Hugging Face incident doesn't happen again. I don't work at a major lab (shout outs to Anima though), and don't have access to any compute independently, so I can't write a paper on this idea or evaluate how well it works in practice. But I'm excited enough about it, and think it's important enough to be trying things like this, that I would be very, very happy if somebody else went and tested

en-gb

10:23

“Do your capabilities homework” by RobinHa

LessWrong (30+ Karma) ·

2026/08/01

It seems to me that a lot of technical ai safety people haven't done their capabilities homework - and that's a shame! I'll try to illuminate here mainly with an example as to why I think people who care about safety should totally pay more attention to the trends and actively engage with them - the case for safe AI not through an additional loss term but as a consequence of the learning algorithm! RLVR It's now been 1.5 years since R1 came out - the paper which really introduced RLVR (RL with

en-gb

33:23

“Why so many therapy etc. frameworks think they’re...

LessWrong (30+ Karma) ·

2026/08/01

If you spend time looking at frameworks in the therapy/meditation/self-help space, you’ll soon find lots of conflicting claims about The One Approach for solving your problems. In the therapy space, there are lots of models that say something like “all emotional problems are caused by trauma/suppressed negative emotions”. Then they disagree about how to deal with it. Internal Family Systems (IFS) therapy holds that you should always move toward the suppressed material cautiously, explicitly

en-gb

4,451 results

LessWrong (30+ Karma)

All Episodes

Similar Podcasts

Life and Technology with Charlie Brown

Tapt Media

Reflections

Livewire Labs

LessWrong (30+ Karma)

LessWrong

Hacker Newsroom

pod pub

UNGOVERNABLE

UNGOVERNABLE