LessWrong (30+ Karma)

LessWrong
  • 0.0 (0)
  • TECHNOLOGY
  • UPDATED DAILY

Audio narrations of LessWrong posts.

  1. 2H AGO

    “Weight-sparse transformers have interpretable circuits” by leogao

    TL;DR: We develop a novel method for finding interpretable circuits in Transformers, by training them to have sparse weights. This results in models that contain very high quality circuits: our circuits are global rather than datapoint dependent; we explain the circuit down to very granular objects, like individual neurons and attention channels, rather than entire MLP layers, attention heads, or groups of nodes; and the circuits are often simple enough to draw in their entirety on a whiteboard. The downside is that our method produces de novo sparse language models, which are extremely expensive to train and deploy, making it unlikely that we will ever be able to use this method to directly pretrain frontier models. We share preliminary results on using sparse models to explain an existing dense model, but our main theory of impact is to eventually scale our method to train a fully interpretable moderate-sized model. If we could fully interpret even (say) a GPT-3 level intelligence, it could aid dramatically in developing a theory of cognition in general. [Blog] [Paper] [Code] Abstract Finding human-understandable circuits in language models is a central goal of the field of mechanistic interpretability. We train models to have more understandable [...] --- First published: November 13th, 2025 Source: https://www.lesswrong.com/posts/yQMQXFAK4mfJjHBpN/weight-sparse-transformers-have-interpretable-circuits --- Narrated by TYPE III AUDIO.

    3 min

  2. 9H AGO

    “What’s so hard about...? A question worth asking” by Ruby

    There's a wide range of tasks that most people get why they’re hard. And then there are activities where I think a lot of people might think to themselves “what's so hard about that?” On the one end of the continuum, you can have a visceral sense of the difficulty of a given task. On the other, you’ve never even given any thought. Things that people appreciate are hard: playing basketball well, solving math puzzles, drawing lifelike pictures, memorizing human anatomy. The commonality here, I think, is that these are the kinds of activities that most people have some experience with. A typical school experience will have you try these all out and you will find that throwing a ball with exactly the right force in exactly the right direction is tricky. Even something that is pretty foreign to people, let's say “rocket science”, might feel hard because people know it involves physics and math, and they’ve tried doing some physics and math. In contrast, there are tasks that, not having tried them, people don’t feel like that should be that hard. That includes me. Until I fired a gun for the first time, I did not appreciate the [...] The original text contained 3 footnotes which were omitted from this narration. --- First published: November 13th, 2025 Source: https://www.lesswrong.com/posts/qnMwwwzmnfaQHvymc/what-s-so-hard-about-a-question-worth-asking --- Narrated by TYPE III AUDIO. --- Images from the article: Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

    5 min

  3. 10H AGO

    “Paranoia rules everything around me” by habryka

    People sometimes make mistakes [citation needed]. The obvious explanation for most of those mistakes is that decision makers do not have access to the information necessary to avoid the mistake, or are not smart/competent enough to think through the consequences of their actions. This predicts that as decision-makers get access to more information, or are replaced with smarter people, their decisions will get better. And this is substantially true! Markets seem more efficient today than they were before the onset of the internet, and in general decision-making across the board has improved on many dimensions. But in many domains, I posit, decision-making has gotten worse, despite access to more information, and despite much larger labor markets, better education, the removal of lead from gasoline, and many other things that should generally cause decision-makers to be more competent and intelligent. There is a lot of variance in decision-making quality that is not well-accounted for by how much information actors have about the problem domain, and how smart they are. I currently believe that the factor that explains most of this remaining variance is "paranoia", in-particular the kind of paranoia that becomes more adaptive as your environment gets [...] --- Outline: (01:31) A market for lemons (05:02) Its lemons all the way down (06:15) Fighter jets and OODA loops (08:23) The first thing you try is to blind yourself (13:37) The second thing you try is to purge the untrustworthy (20:55) The third thing to try is to become unpredictable and vindictive --- First published: November 13th, 2025 Source: https://www.lesswrong.com/posts/yXSKGm4txgbC3gvNs/paranoia-rules-everything-around-me --- Narrated by TYPE III AUDIO.

    23 min

  4. 14H AGO

    “Favorite quotes from ‘High Output Management’” by Nina Panickssery

    Some months ago I read the classic management book High Output Management and made a note of quotes that rang particularly true to me. I normally dislike this genre (management books), and disagree with some popular ones (I sympathize with this review of Scaling People, for example), but found High Output Management pretty reasonable. It's also interesting to see the extent to which its recommendations continue to be followed in successful organizations to this date (the book was published in 1983, but is still popular and recommend amongst tech managers). This post is a list of my copied quotes (headings mine). Delegate activities that are familiar to you Given a choice, should you delegate activities that are familiar to you or those that aren’t? Before answering, consider the following principle: delegation without follow-through is abdication. You can never wash your hands of a task. Even after you delegate it, you are still responsible for its accomplishment, and monitoring the delegated task is the only practical way for you to ensure a result. Monitoring is not meddling, but means checking to make sure an activity is proceeding in line with expectations. Because it is easier to monitor something with [...] --- Outline: (00:46) Delegate activities that are familiar to you (01:38) Should you have personal relationships with your colleagues? (02:08) Use random spot-checks (02:34) On performance reviews (03:47) Assess substance, not potential (05:08) Surprises (05:53) Criticize high achievers (07:18) On interviewing (08:07) Measuring problem-solving ability (09:11) Tricks bad --- First published: November 13th, 2025 Source: https://www.lesswrong.com/posts/jAH4dYhbw3CkpoHz5/favorite-quotes-from-high-output-management --- Narrated by TYPE III AUDIO.

    10 min

  5. 18H AGO

    “The Pope Offers Wisdom” by Zvi

    The Pope is a remarkably wise and helpful man. He offered us some wisdom. Yes, he is generally playing on easy mode by saying straightforwardly true things, but that's meeting the world where it is. You have to start somewhere. Some rejected his teachings. Wisdom Is Offered Two thousand years after Jesus famously got nailed to a cross for suggesting we all be nice to each other for a change, Pope Leo XIV issues a similarly wise suggestion. Pope Leo XIV: Technological innovation can be a form of participation in the divine act of creation. It carries an ethical and spiritual weight, for every design choice expresses a vision of humanity. The Church therefore calls all builders of #AI to cultivate moral discernment as a fundamental part of their work—to develop systems that reflect justice, solidarity, and a genuine reverence for life. The world needs honest and courageous entrepreneurs and communicators who care for the common good. We sometimes hear the saying: “Business is business!” In reality, it is not so. No one is absorbed by an organization to the point of becoming a mere cog or a simple function. Nor can there [...] --- Outline: (00:26) Wisdom Is Offered (02:57) The Context of The Meme Andreessen Used (If You Don't Know) (04:50) Andreessen Takes Bold Stand Against Moral Discernment (07:33) Tech World Decides Performative Cruelty May Have Gone Too Far (11:17) The Avatar of Societal Decay (12:11) Marc's Technical Takes And Arguments Also Are Not Good (13:12) The Best Defense (15:24) You're All Wondering Why You're Here Today --- First published: November 12th, 2025 Source: https://www.lesswrong.com/posts/4gXvnTFy5WCTtYMAA/the-pope-offers-wisdom --- Narrated by TYPE III AUDIO. --- Images from the article: Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

    16 min

  6. 23H AGO

    “Introducing faruvc.org” by jefftk

    I wanted to link an explanation of how far-UVC works, why you might want to use it to clean indoor air, and what we know about its safety. I didn't find anything I liked, so I made something: faruvc.org. You can also link it as far-uvc.org; I got both domains since the hyphenated one reads better but is a bit longer. Let me know if you have ideas for making it better! My goal is to have something anyone can understand, without simplifying so much that it's misleading. At some point I'd like to include an illustration showing far-UVC in use in an occupied space, but I don't have one I like right now. (While I'm an employee of SecureBio, this is a personal project.) Comment via: facebook, mastodon, bluesky --- First published: November 12th, 2025 Source: https://www.lesswrong.com/posts/xpf8ygKPECc8yWhZk/introducing-faruvc-org --- Narrated by TYPE III AUDIO. --- Images from the article: Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

    2 min

  7. 23H AGO

    “Please, Don’t Roll Your Own Metaethics” by Wei Dai

    One day, when I was an interning at the cryptography research department of a large software company, my boss handed me an assignment to break a pseudorandom number generator passed to us for review. Someone in another department invented it and planned to use it in their product, and wanted us to take a look first. This person must have had a lot of political clout or was especially confident in himself, because he refused the standard advice that anything an amateur comes up with is very likely to be insecure and he should instead use one of the established, off the shelf cryptographic algorithms, that have survived extensive cryptanalysis (code breaking) attempts. My boss thought he had to demonstrate the insecurity of the PRNG by coming up with a practical attack (i.e., a way to predict its future output based only on its past output, without knowing the secret key/seed). There were three permanent full time professional cryptographers working in the research department, but none of them specialized in cryptanalysis of symmetric cryptography (which covers such PRNGs) so it might have taken them some time to figure out an attack. My time was obviously less valuable and my [...] The original text contained 1 footnote which was omitted from this narration. --- First published: November 12th, 2025 Source: https://www.lesswrong.com/posts/KCSmZsQzwvBxYNNaT/please-don-t-roll-your-own-metaethics --- Narrated by TYPE III AUDIO.

    4 min

  8. 1D AGO

    “Warning Aliens About the Dangerous AI We Might Create” by James_Miller, avturchin

    Thesis: We should broadcast a warning to potential extraterrestrial listeners that Earth might soon spawn an unfriendly computer superintelligence. Sending the message might benefit humanity. If we were to create an unaligned computer superintelligence, it would likely expand through the universe as quickly as possible. The fastest way would not be by ships, but, as Turchin has described, by sending malicious signals intended for alien civilizations using radio telescopes. Ideally, from the AI's point of view, the signal would be a virus that allows a full download of Earth's AI, which would then take over the receiving civilization. While we still control our radio transmitters, we could send a warning to any alien civilizations that might be listening. The message would explain that Earth may soon undergo a major transition as we develop advanced artificial intelligence, and that some future communications from Earth could be dangerous, and a misaligned superintelligence could send signals containing instructions or code that, if followed, would replicate it and destroy the receiving civilization. The warning should therefore advise aliens never to execute, compile, or build anything based on messages originating from Earth until they have somehow verified that Earth is not controlled by a [...] --- First published: November 12th, 2025 Source: https://www.lesswrong.com/posts/AH4pFd3Sgt6qaoH6h/warning-aliens-about-the-dangerous-ai-we-might-create --- Narrated by TYPE III AUDIO.

    10 min
See All (250)

About

Audio narrations of LessWrong posts.

Information

  • Creator
    LessWrong
  • Years Active
    2023 - 2025
  • Episodes
    250
  • Rating
    Clean
  • Copyright
    © 2025 All rights reserved
  • Show Website
    LessWrong (30+ Karma)

You Might Also Like