Published on January 30, 2026 12:50 PM GMTAnthropic CEO Dario Amodei is back with another extended essay, The Adolescence of Technology.

Published on January 29, 2026 3:48 PM GMTIntroduction to davidad and today's topicstutor valsLessWrong prides itself for an ethos of "say it how you think it" (see "A case for courage when speaking of AI danger").

Published on January 29, 2026 8:24 PM GMTChapter 1: The Prompt

Published on January 29, 2026 2:20 PM GMTUsually the doom conversation is binary. Either the AI will try to take over the world for his own reward function as is discussed in all the alignment problem debates OR a malicious dictator or bad actor will leverage the technology's power to take over the world, create 1984's style of government or many other doom scenarios such as nukes and bio-weapons.

Published on January 30, 2026 6:21 PM GMT1 IntroductionCrosspost of this blog post. My guess is that there will soon be an intelligence explosion.I think the world will witness extremely rapid economic and technological advancement driven by AI progress. I’d put about 60% odds on the kind of growth depicted variously in AI 2027 and Preparing For The Intelligence Explosion (PREPIE), with GDP growth rates well above 10%, and maybe above 100%.

Published on January 29, 2026 10:39 PM GMTA low-effort guide I dashed off in less than an hour, because I got riled up. Try not to hire a team. Try pretty hard at this.Try to find a more efficient way to solve your problem that requires less labor – a smaller-footprint solution.Try to hire contractors to do specific parts that they’re really good at, and who have a well-defined interface.

Published on January 29, 2026 9:00 PM GMTSo, The Possessed Machines. There's been some discussion already. It is a valuable piece -- it has certainly provoked some thought in me! -- but it has some major flaws. It (sneakily!) dismisses specific arguments about AI existential risk and broad swaths of discourse altogether without actually arguing against them.

Published on January 29, 2026 5:09 PM GMTSummary: Current AI systems possess superhuman memory in two forms, parametric knowledge from training and context windows holding hundreds of pages, yet no pathway connects them. Everything learned in-context vanishes when the conversation ends, a computational form of anterograde amnesia. Recent research suggests weight-based continual learning may be closer than commonly assumed.

Published on January 29, 2026 4:33 PM GMT(...but also gets the most important part right.)Bentham’s Bulldog (BB), a prominent EA/philosophy blogger, recently reviewed If Anyone Builds It, Everyone Dies. In my eyes a review is good if it uses sound reasoning and encourages deep thinking on important topics, regardless of whether I agree with the bottom line. Bentham’s Bulldog definitely encourages deep, thoughtful engagement on things that matter.

Published on January 30, 2026 9:21 PM GMTPeople's Clawdbots now have their own AI-only Reddit-like Social Media called Moltbook and they went from 1 agent to 36k+ agents in 72 hours.As Karpathy puts it:What's currently going on at @moltbook is genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently. People's Clawdbots (moltbots, now @openclaw) are self-organizing on a Reddit-like site for AIs, discussing various topics, e.g.

Published on January 30, 2026 9:13 PM GMTMonitoring benchmark/Semi-automated red-teaming for AI controlWe are a team of control researchers supported by CG’s Technical AI safety grant. We are now halfway through our project and would like to get feedback on the following contributions.

Published on January 29, 2026 4:36 PM GMT[W]e’re publishing a new paper that presents the first large-scale analysis of potentially disempowering patterns in real-world conversations with AI.

Published on January 29, 2026 8:42 PM GMTIntroThe eval awareness is a pressing problem, and in this post I advocate that just making benchmarks more realistic won't solve the evaluation awareness problem alone. Better evals must be coupled with other measures, like change in the way models are trained.Safety evaluations are supposed to predict how models will behave in deployment.

Published on January 30, 2026 11:53 PM GMTContext:One way to measure how good LLMs are which is gaining traction and validity is the following:Let T(t) be the time it takes for a human to the task t.

Published on January 30, 2026 9:18 PM GMTtl;dr I wanted to consult the internet for good triple and quadruple entendres. unable to find a good existing compilation, so I wrote my own!

Published on January 30, 2026 11:53 PM GMTContext:One way to measure how good LLMs are which is gaining traction and validity is the following:Let T(t) be the time it takes for a human to the task t.