LessWrong (30+ Karma)

The OpenAI Alignment Research Blog launched today at 11 am PT! With 1 introductory post, and 2 technical posts. Blog: https://alignment.openai.com/ Thread on X: https://x.com/j_asminewang/status/1995569301714325935?t=O5FvxDVP3OqicF-Y4sCtxw&s=19 Speaking purely personally: when I joined the Alignment team at OpenAI in January, I saw there was more safety research than I'd expected. Not to mention interesting thinking on the future of alignment. But that research & thinking didn't really have a place to go, considering it's often too short or informal for the main OpenAI blog, and most OpenAI researchers aren't on LessWrong. I'm hoping the blog is a more informal, lower-friction home than the main blog, and this new avenue of publishing encourages sharing and transparency. --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/tK9waFKEW48exfrXC/announcing-openai-s-alignment-research-blog --- Narrated by TYPE III AUDIO.

For the purposes of this transcript, some high-pitched clicking sounds have been removed. The below is an otherwise unedited transcript of an interview between Dwarkesh Patel[1] and a bat. DWARKESH: Thanks for coming onto the podcast. It's great to have you— BAT: Thanks for having me. Yeah. DWARKESH: You can hear me okay? I mean, uh, all the equip— BAT: Yep, I can hear you. DWARKESH: Great, great. So— BAT: I can hear your voice, too. BAT: If that's what you were asking. DWARKESH: What? No, I was— BAT: Oh, “hear” probably isn’t the right word, I guess. “Sense”? No, it's not “see.” The translation suggestion thing isn’t right. BAT: I can [inaudible] you. It's still so weird to me how humans echolocate through your eyes. DWARKESH: Er, sorry, I was asking— BAT: Yeah, I can also hear your voice. DWARKESH: Uh, great. Okay. DWARKESH: So, the question we’ve all been waiting for, haha: what is it like to be a bat? BAT: Oh, sure. Yeah, that's been everyone's first question. I dunno, what's it like to be a human? Haha. BAT: No, but — I mean, it's not like I’ve ever felt your internal experience. How should I [...] The original text contained 1 footnote which was omitted from this narration. --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/P6BCRfusJefJw6Bei/interview-what-it-s-like-to-be-a-bat --- Narrated by TYPE III AUDIO.

Executive Summary Over the past year, the Google DeepMind mechanistic interpretability team has pivoted to a pragmatic approach to interpretability, as detailed in our accompanying post [[1]] , and are excited for more in the field to embrace pragmatism! In brief, we think that: It is crucial to have empirical feedback on your ultimate goal with good proxy tasks [[2]] . We do not need near-complete understanding to have significant impact. We can perform good focused projects by starting with a theory of change, and good exploratory projects by starting with a robustly useful setting But that's pretty abstract. So how can interpretability help AGI go well? A few theories of change stand out to us: Science of Misalignment: If a model takes a harmful action, we want to be able to rigorously determine whether it was “scheming” or just “confused” [[3]] Empowering Other Areas Of Safety: Interpretability is not a silver bullet that will solve safety by itself, but can significantly help other areas by unblocking things or addressing weak points where appropriate, e.g. suppressing eval awareness, or interpreting what safety techniques taught a model. [...] ---Outline:(00:11) Executive Summary(02:57) Theories Of Change(04:25) Science Of Misalignment(06:59) Empowering Other Areas Of AGI Safety(07:17) Evaluation Awareness(07:53) Better Feedback On Safety Research(08:11) Conceptual Progress On Model Psychology(08:44) Maintaining Monitorability Of Chain Of Thought(09:20) Preventing Egregiously Misaligned Actions(11:20) Directly Helping Align Models(13:30) Research Areas Directed Towards Theories Of Change(13:53) Model Biology(15:48) Helping Direct Model Training(15:55) Monitoring(17:10) Research Areas About Robustly Useful Settings(18:00) Reasoning Model Interpretability(22:42) Automating Interpretability(23:51) Basic Science Of AI Psychology(24:08) Finding Good Proxy Tasks(25:05) Discovering Unusual Behaviours(26:12) Data-Centric Interpretability(27:16) Model Diffing(28:05) Applied Interpretability(29:01) Appendix: Motivating Example For Why Reasoning Model Interp Breaks Standard Techniques {#appendix:-motivating-example-for-why-reasoning-model-interp-breaks-standard-techniques} The original text contained 16 footnotes which were omitted from this narration. --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/MnkeepcGirnJn736j/how-can-interpretability-researchers-help-agi-go-well --- Narrated by TYPE III AUDIO.

Executive Summary The Google DeepMind mechanistic interpretability team has made a strategic pivot over the past year, from ambitious reverse-engineering to a focus on pragmatic interpretability: Trying to directly solve problems on the critical path to AGI going well [[1]] Carefully choosing problems according to our comparative advantage Measuring progress with empirical feedback on proxy tasks We believe that, on the margin, more researchers who share our goals should take a pragmatic approach to interpretability, both in industry and academia, and we call on people to join us Our proposed scope is broad and includes much non-mech interp work, but we see this as the natural approach for mech interp researchers to have impact Specifically, we’ve found that the skills, tools and tastes of mech interp researchers transfer well to important and neglected problems outside “classic” mech interp See our companion piece for more on which research areas and theories of change we think are promising Why pivot now? We think that times have changed. Models are far more capable, bringing new questions within empirical reach We have been [...] ---Outline:(00:10) Executive Summary(03:00) Introduction(03:44) Motivating Example: Steering Against Evaluation Awareness(06:21) Our Core Process(08:20) Which Beliefs Are Load-Bearing?(10:25) Is This Really Mech Interp?(11:27) Our Comparative Advantage(14:57) Why Pivot?(15:20) Whats Changed In AI?(16:08) Reflections On The Fields Progress(18:18) Task Focused: The Importance Of Proxy Tasks(18:52) Case Study: Sparse Autoencoders(21:35) Ensure They Are Good Proxies(23:11) Proxy Tasks Can Be About Understanding(24:49) Types Of Projects: What Drives Research Decisions(25:18) Focused Projects(28:31) Exploratory Projects(28:35) Curiosity Is A Double-Edged Sword(30:56) Starting In A Robustly Useful Setting(34:45) Time-Boxing(36:27) Worked Examples(39:15) Blending The Two: Tentative Proxy Tasks(41:23) What's Your Contribution?(43:08) Jack Lindsey's Approach(45:44) Method Minimalism(46:12) Case Study: Shutdown Resistance(48:28) Try The Easy Methods First(50:02) When Should We Develop New Methods?(51:36) Call To Action(53:04) Acknowledgments(54:02) Appendix: Common Objections(54:08) Aren't You Optimizing For Quick Wins Over Breakthroughs?(56:34) What If AGI Is Fundamentally Different?(57:30) I Care About Scientific Beauty and Making AGI Go Well(58:09) Is This Just Applied Interpretability?(58:44) Are You Saying This Because You Need To Prove Yourself Useful To Google?(59:10) Does This Really Apply To People Outside AGI Companies?(59:40) Aren't You Just Giving Up?(01:00:04) Is Ambitious Reverse-engineering Actually Overcrowded?(01:00:48) Appendix: Defining Mechanistic Interpretability(01:01:44) Moving Toward Mechanistic OR Interpretability The original text contained 47 footnotes which were omitted from this narration. --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/StENzDcD3kpfGJssR/a-pragmatic-vision-for-interpretability --- Narrated by TYPE III AUDIO.

This is a link post. In this paper, we make recommendations for how middle powers may band together through a binding international agreement and achieve the goal of preventing the development of ASI, without assuming initial cooperation by superpowers. You can read the paper here: asi-prevention.com In our previous work Modelling the Geopolitics of AI, we pointed out that middle powers face a precarious predicament in a race to ASI. Lacking the means to seriously compete in the race or unilaterally influence superpowers to halt development, they may need to resort to a strategy we dub “Vassal's Wager”: allying themselves with a superpower and hoping that their sovereignty is respected after the superpower attains a DSA. Of course, this requires superpowers to avert the extinction risks posed by powerful AI systems, something over which middle powers have little or no control over. Thus, we argue that it is in the interest of most middle powers to collectively deter and prevent the development of ASI by any actor, including superpowers. In this paper, we design an international agreement that could enable middle powers to form a coalition capable of achieving this goal. The agreement we propose is complementary to a “verification [...] ---Outline:(01:48) Key Mechanisms(03:36) Path to Adoption(04:48) Urgency --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/rwdtMRmhC9sP9MF5n/how-middle-powers-may-prevent-the-development-of-artificial Linkpost URL:https://asi-prevention.com --- Narrated by TYPE III AUDIO.

Claude Opus 4.5 is the best model currently available. No model since GPT-4 has come close to the level of universal praise that I have seen for Claude Opus 4.5. It is the most intelligent and capable, most aligned and thoughtful model. It is a joy. There are some auxiliary deficits, and areas where other models have specialized, and even with the price cut Opus remains expensive, so it should not be your exclusive model. I do think it should absolutely be your daily driver. Image by Nana Banana Pro, prompt chosen for this purpose by Claude Opus 4.5 Table of Contents It's The Best Model, Sir. Huh, Upgrades. On Your Marks. Anthropic Gives Us Very Particular Hype. Employee Hype. Every Vibe Check. Spontaneous Positive Reactions. Reaction Thread Positive Reactions. Negative Reactions. The Lighter Side. Popularity. You’ve Got Soul. It's The Best Model, Sir Here is the full picture of where we are now (as mostly seen in Friday's post): You want to be using Claude Opus 4.5. That is especially true for coding, or if [...] ---Outline:(00:59) It's The Best Model, Sir(03:18) Huh, Upgrades(04:50) On Your Marks(09:12) Anthropic Gives Us Very Particular Hype(13:35) Employee Hype(15:40) Every Vibe Check(18:16) Spontaneous Positive Reactions(21:44) Reaction Thread Positive Reactions(28:39) Negative Reactions(30:34) The Lighter Side(31:27) Popularity(33:26) You've Got Soul --- First published: December 1st, 2025 Source: https://www.lesswrong.com/posts/HtdrtF5kcpLtWe5dW/claude-opus-4-5-is-the-best-model-available --- 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.

The discussion dives into the impact of modern carbohydrate-heavy diets on insulin resistance. Blood glucose plays a critical role, but too much can cause serious health issues. High glycemic index foods contribute to rapid spikes in hunger and overeating. Historical diets, which had lower glycemic loads, contrast sharply with today's processed options. The conversation also highlights lifestyle contributors like stress and urban planning, linking metabolic dysfunction to our modern environment.

A writer embarks on a daily blogging challenge, committing to post 500 words every day in November. Reflection leads to a candid admission that many posts felt mediocre, influenced by work and vacation distractions. He reveals that several posts were hastily completed from old drafts, abandoning deeper visions. Despite the lower quality, he finds value in clearing out drafts. Ultimately, he concludes that daily posting isn't sustainable without full commitment, weighing reputation against effort and benefits.

The host shares insights from a month of intense writing, completing 30 posts in just 30 days. He discusses the balance between quantity and quality, admitting to lower-effort pieces made to meet daily goals. Managing fatigue becomes a challenge, leading to adjustments in his writing schedule. The conversation also touches on mentoring others and the importance of maintaining commitments. Looking ahead, there’s excitement for future projects while contemplating a return to a more sustainable writing pace.

Ben Goldhaber, a writer and commentator on rationality and self-improvement, shares insights on skill acquisition. He argues that intensive practice over short periods is far more effective than long, spaced efforts. Goldhaber illustrates this with his month-long writing challenge at Inkhaven, emphasizing immersion's benefits, such as varied contexts and rapid skill development. He explores the psychology behind commitment and its impact on self-identity, as well as the reasons many avoid intensive learning despite its benefits.