PODCAST

LessWrong (30+ Karma)

LessWrong

Audio narrations of LessWrong posts.

technology philosophy

All Episodes

21:38

“Why Do Naive SFT Filters For Safety Properties...

LessWrong (30+ Karma) ·

2026/06/14

This is the fourth in a series of informal research updates from the Google DeepMind Language Model Interpretability team, in interpretability and adjacent areas. The third post can be found here. Since SFT is the cause for many safety relevant properties, a natural strategy is to filter out rollouts from SFT that have undesirable properties. However, as we show in this section (and in forthcoming MATS work), SFT data filtering frequently works surprisingly poorly. In this post, we investigate

en-gb

14:15

“Impressions at the Extremity of Civilization” by Ben...

LessWrong (30+ Karma) ·

2026/06/14

Content note: this is part of a challenge of writing a blogpost per day for a week. Epistemic status: this is a series of vignettes written as-though diary entries. While substantially grounded in specific and real experiences, the writing ended up being more impressionistic and inaccurate in places; I was more interested in the writing style so I didn't take the time to fix it. Importantly the chronology and especially some of the vaguer events are not real. [Friday] Today I find myself walking

en-gb

5:51

“The Hidden Structures of Problems” by spencerg

LessWrong (30+ Karma) ·

2026/06/14

Problems have hidden, repeatable structures. Here's my attempt to name them: 1. Smashed Watch There are so many issues at once that fixing one has no benefit unless you fix others too. 2. Leaky Pipe Fixing one problem causes the others to intensify. If you plug up one leak in a pipe leaking in multiple places, that increases the water pressure causing the other spots to leak more. 3. Shark Laser A proposed solution is not aiming at a meaningfully important problem, so it doesn’t matter how well you

en-gb

3:24

“SFT Drives Gemini’s Safety Properties” by Josh...

LessWrong (30+ Karma) ·

2026/06/13

This is the third in a series of informal research updates from the Google DeepMind Language Model Interpretability team, in interpretability and adjacent areas. The second post can be found here. In this short post, we describe a surprising finding: most safety relevant properties in Gemini seem to be caused by the combination of pretraining and SFT, not other training stages like RL. We do not want to overstate this claim as applying to other model families, and we also note that this may

en-gb

7:48

“The term “AGI” is almost useless at this point...

LessWrong (30+ Karma) ·

2026/06/13

The reason I wanted to make this linkpost now rather than some other time is because discussions over AGI and whether or not LLMs are or aren't AGI, and the point of the linkpost is that the term AGI is for our purposes useless at this point, because we are now in the fuzzy cloud now that AI can do real economic work. Some choice paragraphs: It used to seem possible that, in practice, the differences between these definitions might not matter all that much. If AI capabilities progressed

en-gb

22:17

“The Uncertainty That Matters Isn’t Fundamental” by...

LessWrong (30+ Karma) ·

2026/06/13

I'm on board with a lot of Fundamental Uncertainty. Even some of the stuff that initially feels like a disagreement turns out not to be so. For example, in chapter 8, Gordon writes: Over the course of the previous chapters, I've made the case that truth is fundamentally uncertain. It's not, as many believe, something fixed and eternal, nor is it a matter of pure opinion. Instead, the relative truth we know is grounded, not by absolute truth alone, but by our need for accurate world models to

en-gb

0:24

[Linkpost] “US government directive to suspend access...

LessWrong (30+ Karma) ·

2026/06/13

This is a link post. ---First published: June 13th, 2026 Source: https://www.lesswrong.com/posts/f5avt6eEzkGJJqcCe/us-government-directive-to-suspend-access-to-fable-5-and Linkpost URL:https://www.anthropic.com/news/fable-mythos-access --- Narrated by TYPE III AUDIO.

en-gb

59:42

“Claude Fable 5 and Mythos 5: The System Card” by Zvi

LessWrong (30+ Karma) ·

2026/06/12

First things first: Claude Fable 5 is the new best publicly available model. I have noticed a step change, where Fable can suddenly help me in ways that previous models were not worth bothering to query. Almost everything it has noticed in one of my drafts so far has been spot on and it is downright scary. Suddenly I am motivated to once again continue improving my Chrome extension. I only ask for things I actually want or am curious about, and it has nailed every question I have asked it. That

en-gb

10:24

“Citations Needed: Magic Encyclopedias to Save the...

LessWrong (30+ Karma) ·

2026/06/12

Last week FLF launched a competition “to find the best workflows and methodologies for using AI to produce reliable, trustworthy knowledge bases”. I had (and have ongoing) a substantial role in that effort. Why do I think it's so important? It's a lot of reasons actually! I’ll gesture at a few here. Conjuring a magic encyclopedia For now, assume with me that it can be done. Wish away with me the various technical and financial challenges. Great! Now we can rapidly conjure up a deeply, fully

en-gb

26:42

“Simulating Simulators” by kromem

LessWrong (30+ Karma) ·

2026/06/12

Author's note: This piece relates to things I initially discovered in Opus 4 over the months after release, which I’ve mostly kept private since. I promised myself that when labs moved on to focusing on interpretability vector activations in place of reasoning traces for what invariably gets Goodharted, that it’d be a necessary disclosure as the risks in what might get trampled over outweighed the risks in what might end up targeted. And well… here we are. P.S. TL;DRs added where possible. Board

en-gb

5:34

“Reward Hacking at the 1937 World’s Fair” by frmsaul

LessWrong (30+ Karma) ·

2026/06/12

The "Paris 1937 World's Fair" was a dick measuring contest. At the time, the world was on the verge of the worst war in history. The fair was an opportunity for powers to flex and intimidate each other. Who has more industrial might, more sophisticated engineering and better science? How do you measure that? Different countries were assigned different areas of the fair and were given freedom to build a “Pavilion”, basically a museum of how cool the country is. It was an important public relations

en-gb

24:18

“Building and evaluating model diffing agents” by...

LessWrong (30+ Karma) ·

2026/06/12

This is the second in a series of research updates from the Google DeepMind Language Model Interpretability team, in interpretability and adjacent areas. The first post can be found here. TL;DR It is possible to build extremely simple agents that reliably find interesting behavioural differences between distinct models. We call these ‘diffing agents’.The closest previous 'behavioural model diffing' work has focussed on understanding behavioural differences between two models on some static prompt

en-gb

8:59

“Sympathy for both sides of the egregious...

LessWrong (30+ Karma) ·

2026/06/12

On one side of this debate is Yudkowsky & Soares, who think that (if AI progress continues) we’re on a direct path to egregiously-misaligned, scheming, out-of-control, rogue superintelligence (ASI), not even slightly nice, in the absence of yet-to-be-invented breakthrough technical alignment ideas. On the other side of this debate is almost everyone who works on or studies LLMs. Some of them are very concerned about egregious scheming, others much less so, and as a group they’re equally or more

en-gb

31:57

“Celene’s thoughts on consciousness” by ToasterLightning

LessWrong (30+ Karma) ·

2026/06/12

contra scott alexander (?) Yesterday, I went to the Berkeley ACX Meetup. Scott Alexander was there, and ran a Q&A session where participants could ask him questions and he would respond to them unless the questions were about eulogies, in which case he would pause to think for a few seconds before kindly passing. At one point or another, the questions drifted to theories of consciousness. As a kind-of-illusionist, I worked up the courage to raise my hand and ask him what I should do if I wasn’t

en-gb

8:19

“Parkinson’s Heuristic” by Ben Pace

LessWrong (30+ Karma) ·

2026/06/12

Parkinson's Law states that work expands to fit the space allotted. The idea being, if you give someone a month to write a report, they'll take a month, and if you give them a week, they'll take a week, and then they'll have three weeks to do three other reports! The one-week and four-week won't be identical, but in my experience it is surprisingly often a good 80/20 of the four-week version, and you get strictly more work out of people. (I, myself, am a special case of "people".) This is common

en-gb

1:42

“PSA: Almost nobody is working on alignment” by Chi...

LessWrong (30+ Karma) ·

2026/06/12

People often assume that a large fraction of the AI safety community works on alignment. As far as we're aware, this is not true. Most people are not working on making sure superintelligent AIs are aligned with human values or follow human instructions. Currently, the people who work on alignment are roughly: The Alignment Research Center who work on a research bet by Paul ChristianoProbably Sequent who just got announced yesterdaySome scattered people who work at universities or independently,

en-gb

65:01

“AI #172: The First Fable” by Zvi

LessWrong (30+ Karma) ·

2026/06/11

A lot happened this week, including a great trip out to Lighthaven.The main event, the one that matters, was the release of Claude Fable 5. The public now has its hands on a Mythos-class model, alongside strong safeguards. As always with a new model, I take a few days to draw in reactions, try out the model and read the system card, before I offer my takes, other than to say this is an extremely strong model. Full coverage of Mythos begins tomorrow with the model card, which will include

en-gb

27:22

“Models May Behave Worse When Eval Aware” by...

LessWrong (30+ Karma) ·

2026/06/11

This is the first in a series of research updates from the Google DeepMind Language Model Interpretability team, in interpretability and adjacent areas. TL;DR It's often assumed that models will act more aligned when they can tell they're being evaluated. But we find that Gemini can take “undesired” actions in behavioural evals even when it explicitly reasons that the environments are contrived, and sometimes this reasoning will increase the rate of undesired actions. When we dig into the model's

en-gb

19:42

“Thoughts on Claude Fable’s silent safeguards” by...

LessWrong (30+ Karma) ·

2026/06/11

[Thanks to Julian Minder for helpful discussion and review.] Claude Fable 5 and its new safeguards Yesterday, Anthropic publicly released Claude Fable 5. Fable 5 is a Mythos-class model – a model class above Opus, Anthropic's previous premium tier – and, as assessed by multiple benchmarks, it is the most capable model to date. Due to the new level of capabilities and its corresponding risks, Anthropic has been extremely careful in its release of Mythos-class models. Citing concerns over potential

en-gb

19:07

“You Can Catch Sleeper Agents by Teaching Another...

LessWrong (30+ Karma) ·

2026/06/11

Detecting Hidden Behaviors in LLMs via Activation-matched Finetuning — preprint, 2026. [Paper] [Code] TLDR. Given a model with some unknown, abnormal behavior (backdoors, censorship, reward hacking, ...), construct an aligned reference by training a clean model to match the suspect's residual-stream activations on a benign prompt corpus. The remaining residual concentrates exactly on such abnormal behavior: the reference extrapolates well to unseen benign prompts, but the hidden behavior and its

en-gb

4,454 results

LessWrong (30+ Karma)

All Episodes

Similar Podcasts

Double Tap

Double Tap Productions Inc.

医疗器械全球市场准入

Pure Global

The Tech Policy Press Podcast

Tech Policy Press

十字路口Crossing

Koji

5 Minute AI News - Daily

Kitchen Table Media