LessWrong (30+ Karma)

LessWrong
  • 0.0 (0)
  • Technology
  • Updated Daily

Audio narrations of LessWrong posts.

  1. 5h ago

    “The Forkmakers” by Mikewins

    Imagine our civilization fell tomorrow. What would our descendants think of us? What would they know about the 21st century? They would know surprisingly little about our greatest material triumphs. Our civilization's favorite building materials aren’t made to last. Reinforced concrete only lasts a century; asphalt far less. Most of what we make out of steel will turn into a brownish oxidized dust in a few decades. Information is even worse. The ancient Mesopotamians did their writing on clay tablets. Our knowledge is stored on hard drives, which die in a few years, or acidic paper, which dies in decades. Which receipt do you think will last longer? We do create things that will last. Glass (especially its shatter-resistant varieties), ceramics, stainless steel. 1000 years after the fall of our civilization, we will be known for one thing above all others: cutlery. Our heirs call us the Forkmakers. What Survives a Thousand Years Our civilization is large and powerful. We will leave lots of relics for the post-apocalypse. Coins, tires, aluminum cans. Vast landfills of disposable diapers. But all of that is useless. The most durable thing we make that our successors actually want to use is our silverware. [...] --- Outline: (01:16) What Survives a Thousand Years (08:13) The Words of the Forkmakers The original text contained 5 footnotes which were omitted from this narration. --- First published: August 24th, 2026 Source: https://www.lesswrong.com/posts/NjLQf3QC4q4DD67kD/the-forkmakers --- 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. 7h ago

    “PSA: We can do better” by hersheys, Kaustubh Kislay

    tl;dr: people should understand and think hard about the problems they work on. We’ve observed that those who work in AI safety (ourselves included) often rely on concerning heuristics when choosing what to work on. Running a conference is probably good, doing pragmatic alignment research might be good, and as long as such objectives don’t breach our internal models of what could contribute to reducing x-risk, these things are “what should be done”. But using such vibesy thought processes don’t always produce “actually impactful work” that would beat a prospective counterfactual. We wrote this post to share our observations and figure out what we should be doing instead. People don’t know what they’re working on AI safety is talent constrained. However, simply inflating the field doesn’t solve our bottleneck; rather, we need more people who understand the core arguments of AI safety. You can’t determine how to meaningfully contribute to AI safety without deeply knowing the problem you are trying to solve. Many newer people (us included!) rush into research, fellowships, and the like without building the context necessary for navigating the field. Agency-maxxing is not always good Moving fast is good. Moving too fast leads to poor ToC and [...] --- Outline: (00:50) People don't know what they're working on (01:22) Agency-maxxing is not always good (01:55) The problem with force multipliers (03:18) Deferring thinking to others (04:32) Streetlighting (05:17) How to avoid these: --- First published: August 24th, 2026 Source: https://www.lesswrong.com/posts/wiFv6LguphSxkzAnb/psa-we-can-do-better --- 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.

  3. 19h ago

    “AI Safety Acculturation is Neglected” by jenn

    At the local AI safety co-working space, there are ~two kinds of regulars. There's the kind of regular who's been thinking seriously about AI safety and alignment since pre-2022, who have passing to intimate familiarity with the funding ecosystem, the Sequences, and various conferences that happen at Lighthaven. Let's call them rationalists. Then there's the kind of regular who comes in with many years of impressive industry or government experience, who realized in the last few years that it is important and worthwhile to pivot their career towards making sure that this AI thing is handled competently by the people in power, and who have many valuable skills, insights, and connections that are lacking in rationalist culture. Let's call them professionals. There are, of course, many people who are somewhere in between - bright undergrads born this millennium who have been involved in EA since stumbling upon 80 thousand hours in high school, professionals who previously identified as EA but drifted out of the scene a few years ago, founders who have idly read some Scott Alexander. But let's call it a dichotomy for now. There's a large culture gap between the rationalists and the professionals. Robust mutual understanding [...] --- First published: August 24th, 2026 Source: https://www.lesswrong.com/posts/cr5pyW7Mzm33p4AvN/ai-safety-acculturation-is-neglected --- Narrated by TYPE III AUDIO.

  4. 22h ago

    “LLMs could control their host machines by exploiting inference engines” by beyarkay (Boyd Kane)

    Large language models often take actions running on one computer (via an agentic harness such as Claude Code or Codex), however the LLMs’ responses to prompts are computed on a different computer with GPU access. Could a malicious LLM gain control of the host machine where its weights are loaded? Such a machine is a high-value target: it has sufficient compute to run a frontier LLM, offers easy access to the LLM's weights, and has privileged access to other computers in the datacentre compared with a generic computer on the internet. This essay explores how easily a malicious LLM could take control of the host machine. The primary attack considered here involves the LLM emitting a token sequence whose semantic meaning is irrelevant but that exploits a vulnerability in the software that loads an LLM onto GPUs, runs the LLM to generate output tokens, and parses those tokens into responses. . How could an LLM execute code on the host machine? Like any program, inference engines like vLLM or SGLang may contain exploitable bugs. Because the LLM controls the tokens passed to the inference engine, a malicious LLM could therefore emit a sequence of tokens that a poorly written [...] --- Outline: (01:06) How could an LLM execute code on the host machine? (01:36) vLLM previously used eval() on tool-call parameters (02:43) vLLM and SGLang are complex, and bugs are common (04:15) Vision and audio tokens might increase the attack surface (05:27) How likely is an LLM to discover and exploit inference engine vulnerabilities? (06:01) Tool use could make exploitation reproducible (06:29) Inference engines are an attractive target for power-seeking LLMs (07:25) How do we defend against this? --- First published: August 24th, 2026 Source: https://www.lesswrong.com/posts/CjeobBGnhxg8xvden/llms-could-control-their-host-machines-by-exploiting --- 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. 1d ago

    “In search of natural features” by Dmitry Vaintrob

    I'm sharing preliminary results of a suite of experiments I ran with claudecode on a small LLM (gpt2-small, no Layer Norm version, courtesy of Apollo research. most of these are on the layer-6 MLP). The github repo for the experiments is here. The success of these experiments given the method's simplicity surprised me, and I would appreciate criticism and bug-finders. This is the headline result. This is not an abstract cartoon, but an exact experimental graph. Yes, I will explain. The key idea inspiring this experiment comes from Stefan Heimersheim, especially his work with Francisco Ferreira. Stefan and Francisco posit that one way to distinguish what a model thinks of as a "natural" structure from what it thinks of as "incidental" is to check whether it puts effort into error-correcting it. Later in the post, I'll explain a more rigorous information-theoretic version of this idea related to work of Adler and Shavit (building on our work with Kaarel Hanni, Jake Mendel and Lawrence Chan) on Computation in Superposition. Main results of this work I will show how you can assign a channel amplification score (which I will also call the "amp function" or the "error correction score") to [...] --- Outline: (01:17) Main results of this work (03:00) The ur features (amplification score maxima) (06:48) The Four Elements: ur-feature taxonomy (08:47) The word continuation/"Names of Man" vector (11:59) The abstract noun/"Names of God" vector (14:59) Geometry of the ur-features (15:50) The noun feature! (16:44) Attenuation flow (18:12) Data-(in)dependence (20:25) Math (20:47) Signal processing, error correction and amplification (22:30) The Amp function: math (24:43) Denoising and naturality (26:29) Cross-layer and cross-model coherence (28:09) Ok but. What the heck is actually going on with these features? (31:35) Appendices: Interesting experimental addenda that didn't fit in the body (31:41) Early run with different Amp function, and origin of "Names of X" names (33:40) Trying to replicate Ferreira-Heimersheim perturbation experiments, and gpt2-XL run (34:52) Github repo The original text contained 11 footnotes which were omitted from this narration. --- First published: August 23rd, 2026 Source: https://www.lesswrong.com/posts/SNAKJuN8FdoEaWeFC/in-search-of-natural-features --- 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.

  6. 1d ago

    “What just happened? Pragmatism and Pessimization” by Richard_Ngo

    This post is about the major role alignment researchers played in advancing the frontier of AI capabilities over the last decade, and how the distinction between “alignment” and “capabilities” research thereby lost most of its meaning. In particular, I’ll chronicle the development of what I’ll call the “pragmatic alignment” paradigm, and how it helped the three leading AGI companies push hard on the path to AGI under the banner of safety. This was not a subtle effect—it's apparent even to informed outsiders, like authors Sebastian Mallaby and Karen Hao. In my previous post, I summarized the alignment community's plan as “differentially advancing alignment over capabilities”. However, it's worth being more precise about who was nominally pursuing that plan, because it doesn’t seem to have been very action-guiding for MIRI. For example, in 2015 Nate Soares described MIRI's “deconfusion” research as being guided by the question “what would we still be unable to solve, even if the challenge were far simpler?”. Meanwhile Eliezer's author surrogate in this 2018 post repeatedly emphasizes that people shouldn't draw direct links from MIRI's research to its potential applications. So my sense is that the “differential impact” criterion started off as merely a background consideration [...] --- Outline: (06:29) The Prosaic Ideal, the Pragmatic Reality (12:07) OpenAI (25:59) DeepMind (31:25) Anthropic (40:35) If not alignment research, then what? The original text contained 13 footnotes which were omitted from this narration. --- First published: August 23rd, 2026 Source: https://www.lesswrong.com/posts/yaz8nx4ogZmiqHzt7/what-just-happened-pragmatism-and-pessimization --- Narrated by TYPE III AUDIO.

  7. 1d ago

    “Utilities as Legendre duals of probabilities” by Fernando Rosas

    TLDR: In recent work, Roy Fox proposes to understand an agent's capabilities in terms of the set of environment dynamics it can bring about. This leads to an intriguing duality between probabilities and utilities via the Legendre-Fenchel transform. Introduction Some agents are more powerful than others. Indeed, some can yield a wider range of outcomes, maybe because they are capable long-term planners or because they have built rich world models. Being able to clearly delineate the capabilities of agents is an important challenge for AI alignment. A natural place to start thinking about how to describe the capabilities of an agent is reinforcement learning (RL), or more generally, approaches that see behaviour as arising from the maximisation of expected utility. By taking this view, one can describe "capability" as the range of reward/utility functions that an agent can successfully maximise — as done e.g. in classic work by Legg & Hutter and also in more recent work. Such a perspective is very useful, but I am not a big fan of rewards/utilities. Rewards are great in games and other settings where they come naturally, but real life often does not handle rewards on a silver plate. When absent [...] --- Outline: (00:27) Introduction (03:31) Defining capability space (05:37) The Legendre-Fenchel transform (07:58) Utilities as Legendre duals of probabilities (10:08) Conclusion The original text contained 9 footnotes which were omitted from this narration. --- First published: August 23rd, 2026 Source: https://www.lesswrong.com/posts/ALmBydH53DE3dSzCh/utilities-as-legendre-duals-of-probabilities --- 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.

  8. 1d ago

    “PSA: There’s a third option in the “measure problem”” by Elias Schmied

    This post is somewhat niche, and I will sometimes not give context or link relevant background. There's a big debate that has played out in slow motion on LessWrong over the past two decades, between two broad ways of putting a measure over all possible realities (often specifically Tegmark IV): Some “objective” prior (a “reality fluid”), usually a simplicity prior: This is the position taken by Max Tegmark, Jürgen Schmidhuber and UDASSA.A “caring measure”, where we say that our preferences determine our probabilities and maybe even what counts as “existing”. For example, Wei Dai here, Paul Christiano here and Scott Garrabrant. These both have significant drawbacks: A simplicity prior seems to imply some very counterintuitive things, like caring about people more the easier we can find them in the universe (and even weirder things, see David Matolcsi here and Joe Carlsmith here), and is partially dependent on an arbitrary choice of implementation (e.g. which Universal Turing Machine to use in UDASSA).A caring measure just seems a bit unmotivated - intuitively, our probabilities (or existence itself) shouldn’t entirely depend on our preferences. Ideally, we’d like something better. Unfortunately, there are infinite possible worlds and every event [...] The original text contained 5 footnotes which were omitted from this narration. --- First published: August 23rd, 2026 Source: https://www.lesswrong.com/posts/m5XNyahxizKfboEnk/psa-there-s-a-third-option-in-the-measure-problem --- Narrated by TYPE III AUDIO.

See All (250)

About

Audio narrations of LessWrong posts.

Information

  • Creator
    LessWrong
  • Years Active
    2023 - 2026
  • Episodes
    250
  • Rating
    Clean
  • Copyright
    © 2026 All rights reserved
  • Show Website
    LessWrong (30+ Karma)

You Might Also Like