What the smart people are recommending. 8874 books, tools, and products endorsed by the thinkers absorb.md tracks. Ranked by how many times each has been recommended across compiled podcasts, papers, posts, and tweets.
βA new fast confirmation rule mechanism lets you get a hard guarantee that Ethereum will not revert after one slot (12 seconds)β
βMachine learning approaches to spatiotemporal physical systems have primarily focused on next-frame prediction, with the goal of learning an accurate emulator for the system's evolution in time.β
βI'm thrilled that Charlie Bennett and Gilles Brassard have received the A. M. Turing Award for βtheir essential role in establishing the foundations of quantum information science and transforming secβ¦β
βWe present V-JEPA 2.1, a family of self-supervised models that learn dense, high-quality visual representations for both images and videos while retaining strong global scene understanding.β
βMamba-3β
βYou know, one of the stats that people talk about a lot is this METR report which basically says for um each different model that comes out roughly how much human work can it do uh in an automated fasβ¦β
βMore here, including SWE-bench score of 93.9% (!) and a new model behavior known as "answer-thrashing" https://www-cdn.anthropic.com/53566bf5440a10affd749724787c8913a2ae0841.pdfβ
βYou can read a detailed technical report on the software vulnerabilities and exploits discovered by Claude Mythos Preview here: https://red.anthropic.com/2026/mythos-previewβ
βhttps://garryslist.org/posts/sfo-is-trying-to-bury-waymo-4395d272β
βIn this work, we introduce LeWorldModel (LeWM), the first JEPA that trains stably end-to-end from raw pixels using only two loss terms: a next-embedding prediction loss and a regularizer enforcing Gauβ¦β
β@nicoleyh11 offers guidance on handling faculty offers. http://quantumfrontiers.com/2026/03/11/nicoles-guide-to-navigating-faculty-position-offers/β
βBuilding superconducting and neutral atom quantum computersβ
βWeβre sharing new research into how AI might be misused to exploit emotions or manipulate people into making harmful choices.β
βRead more: https://www.anthropic.com/engineering/harness-design-long-running-appsβ
βhttps://garryslist.org/posts/richmond-just-voted-to-reinstate-their-flock-cameras-after-crime-spikedβ
βThis is a very good postβ
βBuilding on earlier work (Meirom, Mor, Weinstein Arxiv 2505.12441), we utilize a benchmarking strategy applicable for comparing these two architectures by measuring "quantumness" directly on optimal sβ¦β
βI'm giving his post to my agent and will ask it to help me talk to it better.β
βRead the full paper: https://transformer-circuits.pub/2026/emotions/index.htmlβ
βIn this paper, we present the AnIML Ontology, an OWL 2 ontology that formalises the semantics of AnIML and aligns it with the Allotrope Data Format to support future cross-system and cross-lab interopβ¦β
βTo read more about standalone functions, check [FUNCTIONS.md](./FUNCTIONS.md).β
βNew report from us: Can you prompt inject your way to an βAβ? ... We tested whether it works. It does on older & smaller models, but not on most frontier AI: https://gail.wharton.upenn.edu/research-anβ¦β
βBuilding and Running Rockefeller Capital Managementβ
βAnd how resilient are you? https://www.mercatus.org/resilient-societies#230548828-4153906252β
βThat's a study by Grid Lab that looks at if Wisconsin hypers scale data centers run on renewables versus gas, what are the cost impacts to all customers?β
βSee Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer for background on this project.β
βRead the paper on arXiv: https://arxiv.org/abs/2603.12230β
β_Blog post by Hartmut Neven_ https://research.google/blog/safeguarding-cryptocurrency-by-disclosing-quantum-vulnerabilities-responsiblyβ
β# Wherefore Art Thou? Provenance-Guided Automatic Online Debugging with Lumosβ
βMy piece in the Economist where I argue against de-weirding AI. It is a strange technology with both risks & opportunities that need to be discovered. Pretending AI works like normal IT automation canβ¦β
βhttps://marginalrevolution.com/marginalrevolution/2026/04/ev-arts-patronage-tranche.html and a hearty thanks to @patrickc for enabling this.β
βI hadn't yet come across Trip's own writeup of the project, which provides way more detail about how he trained the model including SFT (supervised fine tuning) against synthetic chat examples createdβ¦β
βThis paper introduces IDEA2, a novel, semi-automated workflow that integrates Large Language Models (LLMs) within a collaborative, expert-in-the-loop process to address this challenge.β
βThe Hermes Deep Dive. (The new hot AI agent harness).β
βBut you know there are actually demos of of what's called Bzon sampling which was a proposal for using optical uh Ψ§Ω Ψ§Ω uh quantum computing to just demonstrate quantum advantage on a contrived benchmβ¦β
βHere's Anthropic's Mythos system card. The model demonstrated a striking leap in cyber capabilities relative to prior models.β
βPaper review: LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixelsβ
βRead more: https://www.anthropic.com/research/diff-toolβ
βreached by phone shortly before their paper went live, before rushing off to celebrate. https://t.co/4WHulKAPuKβ
βHierarchical Planning with Latent World Modelsβ
βSee the full paper here: https://arxiv.org/abs/2602.11729β
β# The five phases of enterprise AI maturity β and how get past βthe production wallββ
βBut I think a lot about uh this artist Finch wrote this blog post, how to become an artist.β
βMy self-sovereign / local / private / secure LLM setup, April 2026 https://vitalik.eth.limo/general/2026/04/02/secure_llms.htmlβ
βPaper: https://arxiv.org/pdf/2602.10387β
βDiagrams from here: https://arxiv.org/pdf/2509.09677β
βSo with that being said, I want to introduce this benchmark we did CI memories which is a contextual benchmark for um it's a comp competitional benchmark for contextual integrity of persistent memory β¦β
βThere is this very nice paper by Stella from Luther and some other folks um on it's called deep ignorance. So like if you don't want the model to learn something you really just shouldn't put it in thβ¦β
βBig deal paper here: field experiment on 515 startups, half shown case studies of how startups are successfully using AI.β
β# OpenAI Buys TBPN, Tech and the Token Tsunami _Blog post by Ben Thompson_ https://stratechery.com/2026/openai-buys-tbpn-tech-and-the-token-tsunami/β
βLook at the data before spewing nonsense that exposes your ideological brainwashing. https://www.sciencecoalition.org/2025/04/29/sparking-american-economic-growth/#:~:text=Federal%20funding%20for%20baβ¦β
βBenchmark performance is actually limited by token usage. https://joelbkr.substack.com/p/many-benchmarks-scores-would-appear?r=i5f7&utm_medium=ios&triedRedirect=trueβ
βHeterogeneous architectures enable a 138x reduction in physical qubit requirements for fault-tolerant quantum computing under detailed accountingβ
βI think uh maybe Microsoft just released a huge 40-page white paper [music] looking at the jobs that are most likely to be displaced and the ones that are least likely to be displaced.β
βmy friend Sacramento district attorney tenan ho invited me to join him in the audience of the highest court in the land the Supreme Court to hear them discuss exactly these cases Grant's Pass and Martβ¦β
βmy friend Sacramento district attorney tenan ho invited me to join him in the audience of the highest court in the land the Supreme Court to hear them discuss exactly these cases Grant's Pass and Martβ¦β
β# Hong Kong Stocks: Structure & Strategyβ
βThere was actually a very nice Wall Street uh journal article making fun of the controversy. I think it was achieving quantum wokeness is the name if you look it up. It was actually very nice editoriaβ¦β
βThere was actually a very nice Wall Street uh journal article making fun of the controversy. I think it was achieving quantum wokeness is the name if you look it up. It was actually very nice editoriaβ¦β
βSo in fact a year ago we published a paper you know how to build a quantum supercomput which kind of told from the the basic end all the way up to algorithms on how to build everything.β
βThe Toxic Twins Recurrence: Fannie Mae & Freddie Macβ
β# Foundations: U.S. Market Structure & Valueβ
βI have one or two papers with Seth Lloyd on on connections between neural Nets and and Quantum Computing I think it's a very interesting topic but I don't think it has any practical \uh value in the sβ¦β
βFinally he has also written pedagogically explorations of Transformers and the S4 modelβ
βAI Predictions 2025: Comprehensive business forecastingβ
βSOFI: Eleven Questions, Zero Answers (SOFI US)β
βwe have the light paper that folks are probably reading right now and we've released some images of our first chips hereβ
βSo this seminal white paper which was done by a staff software engineer and his team β titled retrieval augmented generation in knowledge graphs for customer service question answering.β
βAnd then this this is another white paper as well from the Microsoft research team in the office of the CTO. So it talks about the benefits of using graph and graph rag over traditional rag for query β¦β
βFinally he has also written pedagogically explorations of Transformers and the S4 modelβ
βthere's this very nice paper published called uh rasp where they developed a basically a programming language that simulates Transformer operationsβ
βthis beautiful paper of uh child's cleave diodo farti gutman and spielman from 20 years ago gave an example of an exponential speed up with a quantum lock and the graph they they consideraβ
βThere is a tutorial on this that is was written and and and and given at a recent conference by one of my colleagues at Fair called Brandon Amos Amos \n very interesting conceptβ
βyou can do a more efficient implementation based on ring lwβ
β# GLM-5.1: Towards Long-Horizon Tasksβ
βyou can modify the basic protocol to incorporate the bell type you know chsh gameβ
βyou can implement trapdoor clause functions based on classical cryptography and in in particular you can base based them on lwe learning with errors which is one of the prime candidates for post quantβ¦β
βWe perform full spectral modeling on the near-peak-light optical and NIR spectra of the SN Ic 2014L to quantitatively constrain helium and other outer-ejecta properties, using the radiative transfer cβ¦β
β_Blog post by Andreessen Horowitz_ https://a16z.com/the-state-of-generative-media-2026β
βIn a wide-ranging Forbes profile, covered by the rundown, Altman said if OpenAI's goal is building AGI that can run companies, his own company should be first in line.β
βRead about our work on warp decode: https://cursor.com/blog/warp-decodeβ
βSo not too long ago the two of you published a paper titled Mathematical Methods and Human Thought in the age of AI. What problems or areas urged this and why now of all times?β
βBut, what I recommend people who want to learn more about it is is read Dario Amodei's two essays.β
βHe's written two really really thoughtful and essays on firstly the machine's loving grace which talks about the aspirations of AI and how much it can opportunities.β
βIntroducing Nested Learning: A new ML paradigm for continual learningβ
β_Blog post by Mustafa Suleyman_β
βThe seminal paper on the meaning of the vector potential by E. J. Konopinski is revisited.β
βThe Next Era of Nuclear Arms Controlβ
βWe talk about that in our blog post \"linear solving for fun and profit.\" Which talks about the the linear solver that manages the sort of construction and and maintenance of this pool of instances iβ¦β
βAnd there's a paper about this called slacker from about 10 years ago \"that\" talks about \"the\" problem that like if you if you look at the like total size of some container image \"it\" may it mayβ¦β
β# What Companies Can Learn from Their Biggest Fansβ
βI wrote a paper called define gravity available out on the internet you can find it.β
βAnd that's that was my mistake because I I mistook β my business for my family and it's like no your f this is not family this is a sports team like we have to be like the championship warriors right β¦β
βIf Primary Care is the Health Systemβs Front Door, Why is it Last in Line for Investment?β
βI'm not sure if you managed to read all 20,000 words of his essay that he put out last week. Uh I think a best good use of Claude would be to put it in there and crunch it down a little bit into the mβ¦β
βthis study was done by Ben Lavine out of UT Southwest and Dallas. And it's a really to me it's just it's seminal groundbreaking study that isn't talked about enough.β
βThe excellent agreement over the intensity range 10 TWcm-2 to 10 PWcm-2 with the recent predictions of A. S. Kornev et al [Phys. Rev. A v.68, art.043414 (2003)] provides the first experimental confirmβ¦β
βWhere AI data centers are reducing power billsβ
βYou can look up situational awareness the decade ahead. It was penned and published online June 2024. And the number of actual hits, predictive hits that Leopold had is staggering.β
βUh one paper recently that uh has come out in this general flavor is from deepseek. This is SPCT self-principled critique tuning.β