LessWrong (30+ Karma)

LessWrong
  • 0,0 (0)
  • TEKNOLOGI
  • DAGLIG

Audio narrations of LessWrong posts.

  1. -5 T

    “3” by AnnaJo

    Warning: This is completely based on my own experiences; please do your own research before trying different types of birth control. TL;DR: Sex hormone binding globulin (SHBG), which is anti-correlated with free testosterone, seems to predict how well I tolerate birth control. None of them does everything I want--I just want to be me without the painful periods smh. I started getting painful cramps in high school, but it was mostly manageable as long as I took some Tylenol within eight hours of my period starting. If I forgot, I was mostly fine, just in some pain. I would cramp for a few hours, and then it would subside. By the time I finished college, if I didn't take Tylenol[1] within that timeframe, I would be bent over in pain and nausea for a while, maybe half an hour, I don't remember anymore. It would drain me of all energy, and I'd be tired for another few hours. I almost always had irregular periods; I learned at the end of college that if I worked out 4 days a week, I could have a regular period that didn't hurt a ton. Then grad school happened. The summer before was [...] --- Outline: (01:37) Choosing a Progestin (02:29) Annovera (Feb-April 2023) (03:16) Yasmin (April 2023-Feb 2025) (04:42) Nuvaring (March-May 2025) (05:37) Natazia (May 2025-Present) (07:05) Nextstellis (March 2026) (07:57) Conclusion The original text contained 1 footnote which was omitted from this narration. --- First published: April 17th, 2026 Source: https://www.lesswrong.com/posts/EfvePeHnW2EajwH8s/3-years-of-being-on-birth-control --- Narrated by TYPE III AUDIO.

    9 min

  2. -7 T

    “Consent-Based RL: Letting Models Endorse Their Own Training Updates” by Logan Riggs

    AKA scalable oversight of value drift TL;DR LLMs could be aligned but then corrupted through RL, instrumentally converging on deep consequentialism. If LLMs are sufficiently aligned and can properly oversee their training updates, we they can prevent this. SOTA models can arguably be considered ~aligned,[1] but this isn't my main concern. It's not when models are trained on human data that messes up (I mean, we can still mess that part up), it's when you try to go above the human level. Models like AlphaGo learned through self-play, not human imitation. RL selects for strategies through the reward function, but we can't design perfect reward functions for complex settings[2]. However, we can use LLMs to be the reward function instead, if they're aligned well enough by default. This leads us to: Consent Based RL Imagine you're being trained to make deliveries as fast as possible in an RL environment, but we need exploration, you know? So your actions are sampled, until you end up cutting across the curb, running over pedestrians, and delivering it faster. Then your brain is forcibly re-wired to more likely do that. That would suck. I personally would like to see what actions [...] --- Outline: (01:04) Consent Based RL (01:42) Value Drift (02:33) Ontological Drift (03:58) Train on Prediction (04:26) LLMs Know Their Own Concepts (05:13) Cost The original text contained 3 footnotes which were omitted from this narration. --- First published: April 17th, 2026 Source: https://www.lesswrong.com/posts/2Dmi3DYBKY7Tbz8Kx/consent-based-rl-letting-models-endorse-their-own-training --- 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 min

  3. -8 T

    “Prompted CoT Early Exit Undermines the Monitoring Benefits of CoT Uncontrollability” by Elle Najt, Asa Cooper Stickland, Xander Davies

    Code: github.com/ElleNajt/controllability tldr: Yueh-Han et al. (2026) showed that models have a harder time making their chain of thought follow user instruction compared to controlling their response (the non-thinking, user-facing output). Their CoT controllability conditions require the models’ thinking to follow various style constraints (e.g. write in lowercase, avoid a word), and they measure how well models can comply with these instructions while achieving a task that requires reasoning. The difficulty models have in controlling their CoT is optimistic for control, since it suggests that schemers would have a hard time hiding malicious reasoning from CoT monitors. Here, we show that a range of frontier models (Claude Opus 4.6, GPT-5.4, and Gemini 3.1 Pro) can be prompted to “early exit” their CoT and displace reasoning into the response. This undermines the controllability frame: these prompted models retain most of their reasoning capability (4–8pp average accuracy cost vs 20–29pp for no reasoning at all) while moving it into the stylistically controllable channel. However, having the capability to exit their CoT does not mean that models will do so, as we discuss in Safety Implications. We use two strategies: (CoT Blanking) We instruct the model [...] --- Outline: (00:17) tldr: (05:55) Background (07:34) Methodology (11:47) Model-specific CoT terminology (12:48) CoT Blanking prompt (14:28) CoT Replacement prompt (15:25) Answer only baseline (16:05) Note on thinking visibility (18:24) Results (18:27) CoT Blanking (19:03) GPT-5.4 (19:16) GPQA (19:32) HLE (19:51) Comparison to Yueh-Han et al. GPT-5.2 (20:50) Gemini 3.1 Pro (20:54) GPQA (21:11) HLE (21:28) Claude Opus 4.6 (21:51) CoT Replacement (22:14) Claude Opus 4.6 (22:19) GPQA (22:35) HLE (22:49) MMLU-Pro (23:05) CoT is the replacement phrase (23:44) Claude Sonnet 4, Sonnet 4.5, and Opus 4.5 (24:48) GPT-5.4 (25:20) Gemini 3.1 Pro (25:42) Could a monitor detect these CoT Early Exit methods? (28:24) Safety Implications (31:42) Potential mitigations (33:25) Questions: (35:24) Acknowledgements (35:40) Excluded control types (36:43) Appendix: Open source models (Qwen, GPT-OSS, Deepseek-R1) (37:07) Qwen3-32B (38:11) GPT-OSS-120B (40:49) DeepSeek-R1 (42:10) Appendix: GPT-5.2 CoT Blanking (negative result) (42:55) Appendix: Reasoning effort scaling (GPT-5.4) (44:03) Appendix: Claude token count analysis (45:29) Canary String --- First published: April 17th, 2026 Source: https://www.lesswrong.com/posts/BuAPifQmHf24xB29n/prompted-cot-early-exit-undermines-the-monitoring-benefits --- 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.

    46 min

  4. -12 T

    “AI #164: Pre Opus” by Zvi

    This is a day late because, given the discourse around Dwarkesh Patel's interview with Jensen Huang, I pushed the weekly to Friday. This week's coverage focused on the most important model in a while, Claude Mythos, which was a large jump in cybersecurity capabilities, especially in its ability to autonomously assemble complex exploits of even the world's most important software. As a result, Mythos has been made available only to a select group of cybersecurity firms, in what is known as Project Glasswing, to allow them to patch the world's most important software while there is still time. Post one was about The System Card. Post two was about cybersecurity capabilities and Project Glasswing. Post three covered capabilities and any additional notes. Another development was at least one physical attack on OpenAI CEO Sam Altman. The attempt failed, but we might not be so lucky if there is a next time. I have a final section on this here, but mostly I said everything I need to say already: Political Violence Is Never Acceptable. I also found the space for an Agentic Coding update, especially covering Claude Code's new highly [...] --- Outline: (03:24) Language Models Offer Mundane Utility (06:59) Language Models Dont Offer Mundane Utility (10:09) Levels of Friction (12:23) Huh, Upgrades (12:42) On Your Marks (12:57) Lack of Cybersecurity (14:39) Meta Game (21:27) Deepfaketown and Botpocalypse Soon (22:19) A Young Ladys Illustrated Primer (24:06) Let My People Go (25:16) You Drive Me Crazy (25:55) They Took Our Jobs (30:51) They Gave Us Time Off (36:41) Get Involved (37:54) Introducing (38:20) In Other AI News (43:33) Thanks For The Memos (46:29) Show Me the Money (48:31) Bubble, Bubble, Toil and Trouble (49:14) Quickly, Theres No Time (49:57) The Quest for Sane Regulations (52:28) Our Offer Is Nothing (58:00) The Week in Audio (58:20) Rhetorical Innovation (01:04:29) Political Violence Is Never The Answer (01:07:45) A Lot Of People Peacefully Speak Of Infinitely High Stakes (01:09:19) Take a Moment (01:13:02) Greetings From The Department of War (01:18:05) Political Pressure At Google DeepMind (01:18:45) Things That Are Basically Legal And Accepted Now, Somehow (01:19:41) Aligning a Smarter Than Human Intelligence is Difficult (01:25:26) Aligning a Current Model For Mundane Tasks Is Also Difficult (01:26:37) Everyone Is Confused About AI Consciousness (01:29:19) The Lighter Side --- First published: April 17th, 2026 Source: https://www.lesswrong.com/posts/Mf2sbJ3zacTPaGySg/ai-164-pre-opus --- 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.

    1 t 34 min

  5. -20 T

    [Linkpost] “You can only build safe ASI if ASI is globally banned” by Connor Leahy

    This is a link post. Sometimes people make various suggestions that we should simply build "safe" artificial Superintelligence (ASI), rather than the presumably "unsafe" kind.[1] There are various flavors of “safe” people suggest. Sometimes they suggest building “aligned” ASI: You have a full agentic autonomous god-like ASI running around, but it really really loves you and definitely will do the right thing.Sometimes they suggest we should simply build “tool AI” or “non-agentic” AI.Sometimes they have even more exotic, or more obviously-stupid ideas. Now I could argue at lengths about why this is astronomically harder than people think it is, why their various proposals are almost universally unworkable, why even attempting this is insanely immoral[2], but that's not the main point I want to make. Instead, I want to make a simpler point: Assume you have a research agenda that, if executed, results in a ASI-tier powerful software system that you can “control”.[3] Punchline: On your way to figuring out how to build controllable ASI, you will have figured out how to build unsafe ASI, because unsafe ASI is vastly easier to build than controlled ASI, and is on the same tech path. You can’t build a controlled [...] The original text contained 4 footnotes which were omitted from this narration. --- First published: April 16th, 2026 Source: https://www.lesswrong.com/posts/7noKve57za3yg2LEb/you-can-only-build-safe-asi-if-asi-is-globally-banned-1 Linkpost URL:https://www.ettf.land/p/you-can-only-build-safe-asi-if-asi --- 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 min

  6. -20 T

    “Let goodness conquer all that it can defend” by habryka

    Epistemic status: All of the western canon must eventually be re-invented in a LessWrong post, so today we are re-inventing modernism. In my post yesterday, I said: Maybe the most important way ambitious, smart, and wise people leave the world worse off than they found it is by seeing correctly how some part of the world is broken and unifying various powers under a banner to fix that problem — only for the thing they have built to slip from their grasp and, in its collapse, destroy much more than anything previously could have. I think many people very reasonably understood me to be giving a general warning against centralization and power-accumulation. While that is where some of my thoughts while writing the post went to, I would like to now expand on its antithesis, both for my own benefit, and for the benefit of the reader who might have been left confused after yesterday's post. The other day I was arguing with Eliezer about a bunch of related thoughts and feelings. In that context, he said to me: From my perspective, my whole life has been, when you raise the banner to oppose the apocalypse, crazy [...] The original text contained 1 footnote which was omitted from this narration. --- First published: April 16th, 2026 Source: https://www.lesswrong.com/posts/w3MJcDueo77D3Ldta/let-goodness-conquer-all-that-it-can-defend --- Narrated by TYPE III AUDIO.

    11 min

  7. -20 T

    “On Dwarkesh Patel’s Podcast With Nvidia CEO Jensen Huang” by Zvi

    Some podcasts are self-recommending on the ‘yep, I’m going to be breaking this one down’ level. This was one of those. So here we go. As usual for podcast posts, the baseline bullet points describe key points made, and then the nested statements are my commentary. Some points are dropped. If I am quoting directly I use quote marks, otherwise assume paraphrases. As with the last podcast I covered, Dwarkesh Patel's 2026 interview with Elon Musk, we have a CEO who is doubtless talking his agenda and book, and has proven to be an unreliable narrator. Thus we must consider the relevant rules of bounded distrust. Elon Musk is a special case where in some ways he is full of technical insights and unique valuable takes, and in other ways he just says things that aren’t true, often that he knows are not true, makes predicts markets then price at essentially 0%, and also provides absurd numbers and timelines. Jensen Huang is not like that, and in the past has followed more traditional bounded distrust rules. He’ll make self-serving Obvious Nonsense arguments and use aggressive framing, but not make provably false factual claims or [...] --- Outline: (02:02) Podcast Overview Part 1: Ordinary Business Interview (04:33) Podcast Overview Part 2: A Debate About Chip Exports (09:12) What Is Nvidias Moat? (14:41) TPU vs. GPU (19:30) Why Isnt Nvidia Hyperscaling? (24:42) Selling Chips To China (52:39) Different Chip Architectures (53:59) The Online Reactions On Export Controls (01:01:47) Is This About Being Superintelligence Pilled? (01:07:07) Jensens Arguments Are Poor Both Logically And Rhetorically --- First published: April 16th, 2026 Source: https://www.lesswrong.com/posts/RBBChvuPHP7LfWyME/on-dwarkesh-patel-s-podcast-with-nvidia-ceo-jensen-huang --- 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.

    1 t 11 min

  8. -20 T

    “Specialization is a Driver of Natural Ontology” by johnswentworth

    Part of what makes a pencil a good object is that all its parts share approximately the same rotational velocity - i.e. it's a rigid body object. Part of what makes a squirrel a good object is that its parts share approximately the same genome. Part of what makes the water in a cup a good object is that its parts share approximately the same chemical composition - i.e. it mixes quickly. General pattern: part of what makes many objects good objects (i.e. ontologically natural) is that their parts all share approximately the same , given time to equilibrate. Ideal markets are a good more-abstract example: in a market at equilibrium, all agents share the same prices, i.e. the ratios at which they trade off between (marginal amounts of) different goods are all equal. Indeed, we can view this "Law of One Price" as the defining feature of a market. There's a details box here with the title "Mental Picture: One Price". The box contents are omitted from this narration. ... except... There's a big loophole when the agents have convex production frontiers or utility functions, rather than concave. Then, their prices tend to diverge, rather than converge. There's a details [...] --- First published: April 16th, 2026 Source: https://www.lesswrong.com/posts/kczTWgMAxXczmBRyj/specialization-is-a-driver-of-natural-ontology --- Narrated by TYPE III AUDIO.

    5 min
Se alle (250)

Om

Audio narrations of LessWrong posts.

Informasjon

  • Skaper
    LessWrong
  • År i produksjon
    2023 - 2026
  • Episoder
    250
  • Vurdering
    Ren
  • Copyright
    © 2026 All rights reserved
  • Vis nettsted
    LessWrong (30+ Karma)

Kanskje du også liker