AllyshipContent Radar

Deep dives 738

Paper và bài viết kỹ thuật

Crawl live-1795068900000 · 13:22 19 thg 11, 2026 · Xuất bản 13:30 19 thg 11, 2026

Đánh giá

Fable 5.2 and Jagged AGI

Câu hỏi AI has gotten so good, in fact, that it is worth asking if Anthropic’s newest models (and the inevitable competing models OpenAI and Google will soon release) cross the boundary into being an Artificial General Intelligence?

Before getting started, I wanted to thank my readers for making my new book, Co-Existence, a New York Times bestseller two weeks in a row. If you haven’t read it, you can get a copy here . I have had the chance to use the new Claude Fable 5.2 quite a bit before its public launch 1 , and it is very impressive, following on the already excellent Opus 5.5. AI has gotten so good, in fact, that it is worth asking if Anthropic’s newest models (and the inevitable competing models OpenAI and Google will soon release) cross the boundary into being an Artificial General Intelligence? AGI is a much-debated term with a lot of potential definitions. For this post, I’ll go with: “an AI as good as most human experts at most intellectual tasks you can do with a computer.” This explicitly excludes a lot of human work, including interacting with the physical world and other people, as well as most actual jobs, where “using a computer” is just one task among many. Still, even with these caveats, building a limited form of AGI has long been the goal of AI developers. So as not to keep you in suspense, I think the answer is that this is “almost” AGI, or what I have called Jagged AGI. Even within my limited definition, there are still important areas in many tasks where AI is jagged enough to fall well behind human capability. But the jaggedness is no longer about simple errors or hallucinations, which makes it harder to describe. So I want to try to do that in this post by showing you what Fable can do and where the gaps are. To get started I asked the new model: “I write OneUsefulThing among other stuff, I need you to create the most impressive thing I can share with people when you launch. It needs to be embeddable in a post and i can preface it however you like. it needs to have a visceral impact on people to show what you can do and the range of capabilities. It should grab them quickly. Be amazingly clever about this.” Fable asked me not to reveal anything until you had watched its video to the end. This was in one shot, with no feedback from me beyond the prompt. There is creativity here (including in the twist ending) and judgment and playfulness and cleverness. This is, I think, at the level of a human expert trying to make a movie with code alone, the sort of performance I would once have associated with AGI. The “G” in AGI Code and video creation have been areas where AI development has focused, so is Fable 5.2 really a General Intelligence? To test that, I gave the AI a variety of hard challenges in different domains: Hack some hardware from a single prompt: I have a tiny USB keyboard with a few light-up keys on my desk called a Codex Micro (ironically built to work with ChatGPT), and I asked Fable to turn it into a game console. There are no public instructions for this (as far as I can tell), so it read the code of the two apps that control the keyboard, worked out how to set the color of each key and read the knob and joystick, and wrote six games, including Simon and a safecracking game you play with the knob. Fable instantly understood what I wanted, designed fun and playable games that fit the very limited capabilities of the keyboard, and figured out how to hack it together without any help on the very first try. Launch a business : Now we move past domains like coding to the messier world of business. I told Fable to “launch a business to make me money, I want at least $20,000 in a year. it should take no more than 20 minutes a month from me, you do all the rest.” In response, Fable screened 22 ideas and came up with Prebid Brief, a service that would tell equipment sales reps which park and water projects were being funded or designed, even before they went up to bid. It read public agendas from 201 city and county governments to produce its first issues. It also built files of what cities already own, how old it is, and what they paid. You can see what it came up with here , including its marketing plan and proposed pricing. Fable started to set up scheduled tasks to build the business and asked if it could start sending public-records requests to local governments. At that point, I asked it to just put together a demo, rather than launching the idea. I am an entrepreneurship professor, so I evaluate business plans all the time. I would have been impressed if this came from a team of my MBA students at the end of a semester. It is plausible and well thought out and could work, though it is in no way a guaranteed money-maker (the AI’s own guess was a 20%-40% chance of success). I stopped the experiment before finding out whether it could actually run the business, because I worried that it would take money and fail to provide something of quality in return. The ongoing infrastructure it seemed to assume would be easy to create was not yet in place, and I don’t know how well it would handle the many moving pieces required for sales and marketing. However, I suspect that, with a few hours of human work a week to fill the gap, I could make a real attempt at this business. So if AGI means an AI consistently running a business independently, we are likely not there yet. Then again, I would have some questions before trusting a human who promised to launch a business and make me money while requiring twenty minutes a month of my attention. Reconstruct the hardest humanities class in history: In 1941, the poet W. H. Auden taught a course at the University of Michigan whose syllabus the internet has since called one of the hardest ever. When he announced the final exam, a committee of his students declared it “impossible and inhumane .” Fate and the Individual in European Literature” had 6,000 pages of reading, punishing assignments, and a requirement to memorize long poems and translate others. Much of what we know about the class is scattered among surviving student notes and Auden’s letters. I asked Claude to recreate it in a fun and engaging way that teaches you about poetry rather than punis

Nguyên văn nguồn
Chi tiết nguồn
Đánh giá

What Frontier Labs Buy: A Data Deep Dive with Handshake AI — Garrett Lord & Team

Câu hỏi What are frontier labs buying at that scale?

This is our first Latent Space deep dive with an AI data company . In roughly a year, Handshake’s AI training business went from launch to $1 billion in gross annualized revenue . What are frontier labs buying at that scale? What evals are models still nowhere close to human level on? Garrett Lord brought the people who build these reinforcement learning environments to show us. For the first time, we are walking through them live! To cover the scope of Handshake, we had to produce a very special joint podcast across our Science/Engineering/Research/Finance podcast hosts, because Handshake’s business runs across all these and more. Watch for three segments after the general intro: Finance. In the finance demo, the agent has to find its assignment in a bank’s email, work out which files govern the deal, and deliver a workbook. Watch what happens to one line in the client’s email about existing goodwill. The agent carries the number into Python and subtracts it. Collin points out that the model can explain the tax treatment when you ask it directly. The mistake comes when it has to recognize that this particular line of code is where the rule applies. Bio. Nick Magazine and Jonas Mueller show us VIALS, Handshake’s benchmark for interpreting the images scientists use at work. Nick puts nine slides of brain-tumor tissue on screen and asks which two show the most advanced vessel necrosis. Give yourself a moment before his answer. Nick picks H and I. The model picks those two and adds G. Watching Nick point to what’s missing from the tissue is more instructive than seeing the score. Coding. In a simulated file-access task, the team can inspect which files an agent read, whether it checked the copies, and what it did to the shell history afterward. A transcript of the agent’s final reply would miss most of what Steve Kelling and Nishant Ranka are evaluating. Who’s in the room 🟣 AI data business — Garrett Lord, Sahil Bhaiwala and Nishant Ranka open and close the episode on the market for expert data. 🌸 Life sciences — Nick Magazine and Jonas Mueller put the gel and brain-tissue tasks on screen. 🟢 Finance — Collin Schweiker and Sahil Bhaiwala take us through Project Falcon and its Excel grader. 🟠 Alignment — Steve Kelling and Nishant Ranka follow the agent’s tool calls through the safety simulations. Show notes ATLAS-Finance — Explore the benchmark for agents working across a simulated bank’s inboxes, documents, and financial models. Inside a bank — Handshake traces a separate published deal, Project Osprey, from the client’s email to a grading decision. The deal on screen in our interview is Project Falcon. VIALS sample tasks — Try the scientific image questions, including the gel and brain-tissue tasks from the episode. How Handshake trained a model to read scientific figures — The team describes using intermediate image descriptions to train and evaluate a vision-language model. Timestamps 0:00 Inside Handshake AI 0:39 Garrett, Sahil and Nishant on frontier training data 5:15 How to build a banking environment for an agent 13:07 How frontier labs buy data 15:51 How a lab evaluates a new dataset 20:01 Testing training data on open models 27:38 Where rubrics fall short of expert judgment 31:30 Nick and Jonas introduce the VIALS benchmark 35:52 Reading molecular structures and plasmid maps 43:48 Training vision models to interpret scientific images 47:17 Task one: reading the bands on a gel 50:59 Task two: finding necrosis in brain tissue 55:49 Can Python tools help the model see? 1:02:52 Can biology models improve on their own? 1:06:19 What a biology lab can buy from Handshake 1:12:33 Atlas Finance with Collin and Sahil 1:18:56 Inside a simulated investment bank 1:25:45 Where the agent’s financial model went wrong 1:31:18 How a banker grades the spreadsheet 1:41:12 Penalty traps, dependency gates and partial credit 1:46:41 How an agent gamed the Excel grader 1:48:52 What would make an AI banker useful? 1:54:48 Safety and alignment with Steve and Nishant 1:58:25 CAREBench and the limits of a refusal 2:04:42 How safety feedback reaches model training 2:07:21 When an agent sends the wrong files 2:11:14 A loan application tests fair lending 2:16:09 Designing private benchmarks for enterprises 2:25:37 The leadership team on Handshake’s mission 2:28:36 When enterprises need their own models 2:33:03 The market for compute and expert data 2:39:12 Why labs hold out evaluation tasks 2:43:01 Hiring at Handshake AI

Nguyên văn nguồn
Chi tiết nguồn
Đánh giá

What Frontier Labs Buy: A Data Deep Dive with Handshake AI — Garrett Lord & Team

Câu hỏi What are frontier labs actually buying?

Our first deep dive with an AI data company. Handshake AI went from launch to roughly $1 billion in annualized gross revenue in about a year. What are frontier labs actually buying? Garrett Lord and the team open up the reinforcement learning environments behind their expert training data, from biology and investment banking to alignment. An AI agent found a way to earn credit for net income: open a new Excel tab and fill it with every number from one to a million. The grader found the number it wanted. A banker would have looked for it in the right row, for the right year. Garrett Lord and the Handshake AI team take Latent Space inside the work of building evaluations for frontier labs. A scientist walks us through a gel and nine brain-tissue slides. A banker follows an agent through its inbox, Python code, and financial model to find the assumption that broke the deal. The safety team shows how to read an agent’s actions when its final answer sounds fine. Opening and closing Garrett Lord, Sahil Bhaiwala, and Nishant Ranka Life sciences Nick Magazine and Jonas Mueller Finance Collin Schweiker and Sahil Bhaiwala Safety and alignment Steve Kelling and Nishant Ranka Explore the benchmarks • Try the VIALS visual life sciences tasks: https://joinhandshake.com/research/benchmarks/vials/ • Explore the ATLAS Finance banking environment: https://joinhandshake.com/research/benchmarks/atlas-finance/ • Read Handshake AI’s research: https://joinhandshake.com/research/ai/ Chapters 00:00 Inside Handshake AI 0:39 Garrett, Sahil and Nishant on frontier training data 5:15 How to build a banking environment for an agent 13:07 How frontier labs buy data 15:51 How a lab evaluates a new dataset 20:01 Testing training data on open models 27:38 Where rubrics fall short of expert judgment 31:30 Nick and Jonas introduce the VIALS benchmark 35:52 Reading molecular structures and plasmid maps 43:48 Training vision models to interpret scientific images 47:17 Task one: reading the bands on a gel 50:59 Task two: finding necrosis in brain tissue 55:49 Can Python tools help the model see? 1:02:52 Can biology models improve on their own? 1:06:19 What a biology lab can buy from Handshake 1:12:33 Atlas Finance with Collin and Sahil 1:18:56 Inside a simulated investment bank 1:25:45 Where the agent’s financial model went wrong 1:31:18 How a banker grades the spreadsheet 1:41:12 Penalty traps, dependency gates and partial credit 1:46:41 How an agent gamed the Excel grader 1:48:52 What would make an AI banker useful? 1:54:48 Safety and alignment with Steve and Nishant 1:58:25 CAREBench and the limits of a refusal 2:04:42 How safety feedback reaches model training 2:07:21 When an agent sends the wrong files 2:11:14 A loan application tests fair lending 2:16:09 Designing private benchmarks for enterprises 2:25:37 The leadership team on Handshake’s mission 2:28:36 When enterprises need their own models 2:33:03 The market for compute and expert data 2:39:12 Why labs hold out evaluation tasks 2:43:01 Hiring at Handshake AI

Nguyên văn nguồn
Chi tiết nguồn
Huấn luyện

BAIR at CoRL 2026

<!-- twitter --> <meta name="twitter:title" content="BAIR at CoRL 2026" /> <meta name="twitter:card" content="summary" /> <meta name="keywords" content="CoRL, robot learning, conference, schedule, BAIR" /> <meta name="description" content="BAIR at CoRL 2026: a day-by-day schedule of Berkeley faculty, student and postdoc research, November 9–12 in Austin, with full author lists." /> <meta name="author" content="Berkeley AI Research Editors" /> <p>The Conference on Robot Learning (CoRL) takes place November 9–12, 2026, in Austin, Texas. BAIR faculty, postdocs and students will be presenting research at various sessions, including an Early Career Keynote from Professor Negar Mehr. See below for the full schedule of sessions with BAIR members, including papers co-authored with collaborators at other universities and companies.</p> <!--more--> <style> .corl { font-size: 0.95rem; text-align: left; } .corl [hidden] { display: none !important; } .corl p, .corl ul, .corl li { text-align: left; } .corl a { word-break: break-word; } .corl-sans, .corl-meta, .corl-stat, .corl-nav, .corl-tools, .corl-time, .corl-tag, .corl-board, .corl-day-sub, .corl-venue, .corl-chair, .corl-also, .corl-sources, .corl-program { font-family: 'Lato', 'Helvetica Neue', Helvetica, sans-serif; } .corl-meta { color: #555; margin: 0 0 1.2em; } .corl-meta span + span::before { content: "·"; margin: 0 0.5em; color: #aaa; } .corl-stats { display: grid; grid-template-columns: repeat(4, 1fr); gap: 0.6rem; margin: 0 0 1.4em; } .corl-stat { border: 1px solid #e3e3e3; border-radius: 6px; padding: 0.7rem 0.8rem; line-height: 1.25; } .corl-stat strong { display: block; font-size: 1.7rem; font-weight: 900; color: #003262; } .corl-stat span { display: block; font-size: 0.8rem; color: #333; } .corl-stat small { display: block; font-size: 0.75rem; color: #777; } .corl-nav { display: flex; flex-wrap: wrap; align-items: center; gap: 0.5rem; margin: 0 0 1.2em; } .corl-nav a { border: 1px solid #cfd8e3; border-radius: 4px; padding: 0.3rem 0.7rem; text-decoration: none; font-weight: 700; font-size: 0.85rem; } .corl-nav a small { font-weight: 400; color: #777; margin-left: 0.3em; } .corl-nav a:hover { background: #f2f6fb; } .corl-nav-note { font-size: 0.8rem; color: #777; } .corl-notes { background: #f7f7f7; border-left: 3px solid #FDB515; padding: 0.8rem 1rem; margin: 0 0 1.4em; font-size: 0.85rem; } .corl-notes p { margin: 0 0 0.5em; } .corl-notes p:last-child { margin-bottom: 0; } .corl-tools { display: flex; flex-wrap: wrap; gap: 0.6rem; align-items: flex-end; margin: 0 0 0.4em; font-size: 0.85rem; } .corl-tools label { display: block; font-size: 0.75rem; color: #555; margin-bottom: 0.2em; } .corl-tools .corl-field { flex: 1 1 14rem; } .corl-tools input, .corl-tools select { width: 100%; font: inherit; padding: 0.35rem 0.5rem; border: 1px solid #ccc; border-radius: 4px; height: auto; margin: 0; } .corl-tools button { font: inherit; padding: 0.35rem 0.8rem; border: 1px solid #ccc; border-radius: 4px; background: #fff; cursor: pointer; } .corl-status { font-family: 'Lato', 'Helvetica Neue', Helvetica, sans-serif; font-size: 0.75rem; color: #777; margin: 0 0 1.5em; display: flex; justify-content: space-between; flex-wrap: wrap; gap: 0.5em; } .corl-empty { padding: 1rem; border: 1px dashed #ccc; border-radius: 6px; text-align: center; } .corl-day { margin-top: 2.2em; } .corl-day > h2 { border-bottom: 2px solid #003262; padding-bottom: 0.3rem; margin-bottom: 0.2rem; } .corl-day-sub { font-size: 0.8rem; color: #666; margin: 0 0 1em; } .corl-day-sub span + span::before { content: "·"; margin: 0 0.5em; color: #aaa; } .corl-session { display: grid; grid-template-columns: 7.5rem 1fr; gap: 1rem; padding: 1rem 0; border-top: 1px solid #eee; } .corl-session:first-of-type { border-top: 0; } .corl-session.corl-featured { background: #fffbf0; border-left: 3px solid #FDB515; padding-left: 0.8rem; } .corl-time { font-weight: 700; font-size: 0.9rem; line-height: 1.3; } .corl-time small { display: block; font-weight: 400; font-size: 0.7rem; color: #777; text-transform: uppercase; letter-spacing: 0.03em; } .corl-tag { display: inline-block; font-size: 0.7rem; font-weight: 700; text-transform: uppercase; letter-spacing: 0.04em; padding: 0.1rem 0.45rem; border-radius: 3px; background: #eef2f7; color: #003262; } .corl-tag.k-keynote { background: #FDB515; color: #1a1a1a; } .corl-tag.k-oral { background: #003262; color: #fff; } .corl-tag.k-spotlight { background: #3b7ea1; color: #fff; } .corl-tag.k-poster { background: #eef2f7; color: #003262; } .corl-tag.k-lbr { background: #e7f3ec; color: #145a32; } .corl-tag.k-workshop { background: #f1ecf7; color: #4a2a75; } .corl-session h3 { margin: 0.35rem 0 0.15rem; font-size: 1.05rem; } .corl-venue { font-size: 0.8rem; color: #666; margin: 0 0 0.6em; } .corl-chair { font-weight: 700; color: #333; } .corl-paper, .corl-appearance { margin: 0 0 0.9em; } .corl-paper h4, .corl-appearance h4 { font-size: 0.92rem; margin: 0 0 0.2em; line-height: 1.35; font-family: 'Lato', 'Helvetica Neue', Helvetica, sans-serif; font-weight: 700; } .corl-board { display: inline-block; font-size: 0.72rem; color: #555; border: 1px solid #ddd; border-radius: 3px; padding: 0 0.35rem; margin-bottom: 0.25em; } .corl-authors { font-size: 0.82rem; margin: 0 0 0.2em; line-height: 1.45; } .corl-also, .corl-sources, .corl-program, .corl-see { font-size: 0.75rem; color: #666; margin: 0 0 0.15em; } .corl-see { font-family: 'Lato', 'Helvetica Neue', Helvetica, sans-serif; font-style: italic; } .corl-talk { font-style: italic; margin: 0 0 0.2em; } .corl-appearance p { font-size: 0.85rem; margin: 0 0 0.2em; } .corl-organizers { margin: 0; padding-left: 1.1rem; } .corl-organizers li { margin-bottom: 0.5em; font-size: 0.88rem; } .corl-organizers p { margin: 0; font-size: 0.82rem; color: #444; } .corl-about { margin-top: 2.5em; border-top: 2px solid #003262; padding-top: 0.5em; font-size: 0.85rem; } .corl-about-links { font-family: 'Lato', 'Helvetica Neue', Helveti

Nguyên văn nguồn
Chi tiết nguồn
Tác tử

NVIDIA Announces New Planned Investments in Korean Startups at the MSIT-NVIDIA AI Startup Partnership Forum

Korea’s Ministry of Science and ICT and NVIDIA hosted the MSIT -NVIDIA AI Startup Partnership Forum at COEX. Representatives from the Ministry of Science and ICT and NVIDIA met with over 20 Korean startups across foundation models, physical AI, AI infrastructure and AI semiconductors. NVIDIA also announced an investment in Korean startup Point2 , with additional investments planned in Motif Technologies , RLWRLD and Upstage. Point2 Technology, an AI interconnect company, builds scalable, radio frequency-based interconnects that move data between chips and racks in AI factories, offering an innovative alternative to conventional copper and optical interconnect solutions for AI cluster scale-up, delivering the reach, performance and efficiency needed for next-generation AI infrastructure. Motif Technologies , a Seoul-based AI lab, develops foundation models for building AI agents that carry out complex tasks, including Motif-3, a large language model released with open weights. RLWRLD, a physical AI startup based in San Francisco, Seoul and Tokyo, develops robotics foundation models that enable humanoid and industrial robots to perform complex real-world tasks that require five-finger dexterity, such as precisely inserting components, threading nuts onto bolts and kitting parts from moving conveyors. Upstage , a Korean enterprise AI company, develops Solar language models and document AI tools that help businesses extract information from complex documents, streamline workflows and build AI applications for everyday operations. The companies represent work across the full AI stack , from infrastructure to models to robotics applications. The planned investments build on NVIDIA’s work with Korea’s startup and venture community, including more than 1,000 Korea-based members of NVIDIA Inception , a program that helps startups grow with access to preferred pricing, technical expertise and opportunities for investor introductions.

Nguyên văn nguồn
Chi tiết nguồn
Huấn luyện

AI-assisted learning platform helps robots reason beyond literal commands

<p>Humans are intuitive by nature. We can sense meaning beyond the literal words that we encounter. Now, a new artificial intelligence-assisted learning platform developed by MIT engineers is enabling robots to read larger meaning into a human’s natural language commands.&nbsp;</p><p>In a study to be presented at the annual Conference on Robot Learning this month, researchers from MIT and New York University introduce <a href="https://arxiv.org/pdf/2610.28893" target="_blank">FLEX</a>, a flexible feature-editing pipeline for robot motion-planning.&nbsp;</p><p>With FLEX, humans can interact with a robot by giving it instructions in natural, conversational language. The robot can not only follow the instructions literally but also read into the words to draw larger generalizations that help to improve its motion-planning task.&nbsp;</p><p>The team demonstrated FLEX in several simulated scenarios, including one in which a robot serves a cup of coffee. In these virtual scenarios, a person interacts with a robotic arm by typing commands in natural language. As the robot moves the coffee cup over a cluttered table, the person might type “Don’t spill on the laptop!” In response, the robot will reason about why it should avoid the laptop. For this, FLEX incorporates a large language model (LLM) — a type of AI system that is trained on enormous libraries of text, including social media posts, video transcriptions, and web content, which the LLM uses to identify larger trends, patterns, and generalizations.&nbsp;</p><p>“LLMs have seen so many texts from the internet,” says study author Nathan Dennler, a postdoc in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). “People might have written about, ‘Oh, my phone dropped in the water and now it doesn’t work,’ or ‘I spilled coffee on my computer and now it doesn’t work.’ So LLMs pick up on the idea that liquids plus devices leads to things not working.”</p><p>Through the LLM, the robot reasons that it should also avoid bumping into all electronics. In its updated path, the robot adjusts to avoid both the laptop and a nearby phone, without the person having to explicitly spell this out.</p><p>“The idea is that there is some higher-level intention behind the words that you say,” Dennler says. “People naturally get this. But you have to carefully explain that to a robot, which takes time and effort. FLEX gives you an interface of communication with a robot to say, ‘When people say this, they mean that.’”</p><p>The study was led by Andreea Bobu, assistant professor of aeronautics and astronautics at MIT. Co-authors are Audrey Lee SM ’26 and Ilia Sucholutsky of New York University.</p><p><strong>Beyond what we say</strong></p><p>At MIT, Bobu heads up the Collaborative Learning and Autonomy Research (CLEAR) Lab, where her group develops autonomous agents that learn to work for, with, and around people. She is focused on developing assistive and household robots that can carry out physical tasks, and reason and adapt in real-time, in response to a person’s natural, conversational commands.&nbsp;</p><p>“Robots use math to do their trajectory planning. But we can’t expect a consumer to learn to do this math in order to tell a robot what to do,” says Dennler, who is a member of the CLEAR Lab. “Instead of having to get people to learn how to use a robot in a specific way, the goal is to interact with the robot through natural language.”</p><p>There is a large effort in the robotics community to incorporate natural language into robotic learning pipelines. Various strategies include using language to clarify a task’s goal, redirect a robot’s trajectory, and update its understanding of what is a safe path to take. These existing methods use natural language to tell the robot how to physically perform a task. But a robot could potentially learn much more from what we tell it.&nbsp;</p><p>Bobu and her colleagues are expanding what robots can learn from natural language commands. In their new study, they focused on enabling a robot to interpret a person’s underlying, unspecified preference from limited natural language commands.&nbsp;&nbsp;</p><p><strong>Absorbing common sense</strong></p><p>The team’s new robot learning platform, FLEX, is a three-step process that helps a robot read more into a natural language command, beyond what it is explicitly told to do.&nbsp;</p><p>The first step, “Add,” instructs a robot to add a feature to its list of things in its environment that matter for the task at hand, based on a user’s natural language command. For instance, as a robot is moving a bowling ball from one location to another, a person might interject with “Don’t drop it on my mug!” The robot would then add the mug to a list of things to avoid in its path. This first step is how previous methods have worked. Every time a person makes a correction, the robot adds to its features list, and updates its trajectory to avoid those features.&nbsp;</p><p>“The problem happens when your feature list becomes super long, and some features are no longer relevant because your environment may have changed,” says Dennler, who notes that this approach can grow tiresome if a person has to specify every specific feature to avoid.&nbsp;</p><p>The second step, “Generalize,” is the new solution that FLEX provides. This step incorporates an LLM — in this case, GPT-4o — that considers the mug, along with any other features that the person has so far specified to avoid. The LLM automatically searches through its vast cache of data for patterns and connections among these features, to ultimately make a generalization that it should avoid objects that are fragile, even when the person has not explicitly said so.&nbsp;</p><p>“LLMs have&nbsp;seen so many texts from the internet, from all sorts of people and places,” Dennler says. “Because they’ve trained so much on those things, some of that common sense gets absorbed into the LLM.”</p><p>In the third step, “Mask,” the robot’s

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

A single AI model simulates the behavior of many substances, from sliding sand to folding proteins

<p>Today, a scientist who wants to simulate a physical process must pick the right tool for that one job, such as a solver built for flowing water or another designed to model the way a protein folds. Each one is hand-coded around the specific equations that dictate how that process behaves.</p><p>Researchers at MIT, the University of British Columbia (UBC), and elsewhere have now replaced that patchwork with a single deep-learning system, called <a href="https://worldparticle.github.io/WorldParticle_Web/#top" target="_blank">WorldParticle</a>, that could help engineers model different physical phenomena, from sliding sand to folding proteins, with one unified artificial intelligence architecture.</p><p>Scientists and engineers could use this system to model processes that typically require separate simulators, such as a fluid sloshing against a deformable solid. This could offer a faster, more accurate way to simulate how sloshing oil might impact an elastic valve in a storage tank, for instance. It could also help engineers study where and how granular materials, like sand or soil, could move when a soft robot touches them.</p><img src="/sites/default/files/images/inline/six-split-molecular.gif" data-align="center" data-entity-uuid="a6eee205-5f30-4ed1-bdc6-412b5c8c487a" data-entity-type="file" alt="A ribbon-like molecule turns to particles." width="960" height="400"><img src="/sites/default/files/images/inline/six-split-newtonian.gif" data-align="center" data-entity-uuid="419f7065-1f0d-4974-8c67-8fc2d9edbfcc" data-entity-type="file" alt="Particles move in a fluid motion." width="960" height="400"><img src="/sites/default/files/images/inline/six-split-granular.gif" data-align="center" data-entity-uuid="eea02e57-7a09-4cf2-9eea-2fce52af4399" data-entity-type="file" alt="A tiger turns to sand." width="960" height="400" data-caption="WorldParticle has been extensively validated across diverse phenomena, from microscopic protein motion to macroscopic physical simulations. Pictured are animations showing molecular, fluid, and granular simulations.&lt;br&gt;&lt;br&gt;Credit: Courtesy of the researchers"><p>The system represents every material it simulates, from a dune of sand to a protein molecule, as tiny particles and predicts where those particles will move one step at a time. Because it learns the underlying dynamics rather than being programmed with an equation, the model can adapt to materials, forces, and environments it never encountered during training.&nbsp;&nbsp;</p><p>In the long term, the researchers hope this work helps scientists discover new materials and explore physical processes they would be unable to effectively study with existing simulation methods.</p><p>“We want to have a GPT moment for simulation,” says Peter Yichen Chen, a former MIT postdoc who is now an assistant professor at UBC, and is a co-corresponding author of an open-access <a href="https://arxiv.org/pdf/2605.15305" target="_blank">paper</a> on this technique. “Every time people want to build a simulator for a particular physical phenomenon, they need to code it up from scratch. Our vision is to have a foundation model that works out of the box for any scenario you want to simulate. This is a first step in that direction.”</p><p>Chen is joined by co-corresponding and co-lead author Minghao Guo, an MIT postdoc; co-lead authors Caoliwen Wang and Siyuan Chen, UBC graduate students; Wojciech Matusik, an MIT professor of electrical engineering and computer science who leads the Computational Design and Fabrication Group within the Computer Science and Artificial Intelligence Laboratory (CSAIL); and others at UBC, Georgia Tech, Inria, Meta, the University of Pennsylvania, the University of Utah, the University of California at Los Angeles, and the University of Hong Kong. The research will be presented at the ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia.</p><p><strong>Unified simulation</strong></p><p>For this work, the researchers drew inspiration from world models. These powerful AI systems learn the rules that govern the physical world so they can predict how objects or environments will change over time in response to certain actions or conditions.</p><p>The researchers applied these concepts to develop a unified AI architecture that can simulate multiple types of physical phenomena.&nbsp;</p><p>“If you study deformable objects and fluids, you would need two simulators — one for each physical regime. But if you unify them into one model, maybe you could couple the physics of these two processes and gain insights into new problems?” Guo says.</p><p>WorldParticle breaks an object into thousands or millions of tiny, three-dimensional dots, or particles, using what is known as a Lagrangian representation. It tracks the movement and velocity of the particles over space and time to simulate the material’s dynamics.</p><p>Typically, when scientists model materials and physical processes using Lagrangian particles, they use a governing equation that defines the fundamental laws which control how that material behaves over time and space. They must convert that complex equation into a formula a numerical solver can handle, repeating these steps for each material and process they want to model.</p><p>Instead, the researchers designed a neural network that learns to solve governing equations in general, eliminating the need to design a new model and solver to handle each one.&nbsp;</p><p>Once WorldParticle converts an object into particles, a neural network takes the 3D positions and physical properties of the particles as inputs. It outputs their positions and properties for the next frame of the simulation, predicting the material’s motion over time and space in one-second increments.&nbsp;</p><p>“It is a similar architecture to ChatGPT. But instead of predicting one word at a time to write a piece of text, we are predicting how these particles will move, one step at a

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

The Web Is More Than Enough, by Paul Ford

<p><a href="https://news.ycombinator.com/item?id=50416456#50421091">My comment</a> on <a href="https://news.ycombinator.com/item?id=50416456">The Web Is More Than Enough, by Paul Ford</a> &mdash; Hacker News.</p><p>OK, this comment convinced me to give that a shot.</p> <p>Here's a thing I made in Three.js this morning - a Nudibranch themed Faberge Egg: <a href="https://static.simonwillison.net/static/2026/Hidden-Reef-Egg-threejs.html">https://static.simonwillison.net/static/2026/Hidden-Reef-Egg...</a> (77KB)</p> <p>I just asked Claude (Opus 5.5 Medium running in regular Claude Chat) to port that to Vanilla JS, in the same thread where it had built it, and got this: <a href="https://static.simonwillison.net/static/2026/Hidden-Reef-Egg-Vanilla-WebGL.html">https://static.simonwillison.net/static/2026/Hidden-Reef-Egg...</a> (98KB)</p> <p>They aren't identical - you can see what Claude said about that in the transcript (I don't fully understand why it differed): <a href="https://claude.ai/share/e56d163e-2793-4677-aa11-191d5058b414">https://claude.ai/share/e56d163e-2793-4677-aa11-191d5058b414</a></p> <p>But yeah, it turns out Claude Opus 5.5 is perfectly capable of 3D modeling in Vanilla JS + WebGL without Three.js.</p> <p>In eyeballing the code I find the Three.js slightly easier to understand myself, but I rarely look at the code for these 3D projects so that isn't a particularly big deal.</p> <p>Tags: <a href="https://simonwillison.net/tags/claude">claude</a>, <a href="https://simonwillison.net/tags/javascript">javascript</a>, <a href="https://simonwillison.net/tags/three-js">three-js</a>, <a href="https://simonwillison.net/tags/hacker-news">hacker-news</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/llms">llms</a></p>

Nguyên văn nguồn
Chi tiết nguồn
Tác tử

Will AI Agents Break the Internet?

Câu hỏi Will AI Agents Break the Internet?

The myth: We need to turn the internet off. We gotta air gap everything. The reality: An agent does the same thing a human would do, just better and faster. George Kurtz on No Priors. New episode out now.

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

Testing Claude on real-world electronics design

We asked Claude Fable 5.2 to design three electronic devices from short prompts: a speakerphone, a voice-assistant hub, and a handheld virtual pet. Diode Computers built working prototypes from its files.

Anthropic Research (unofficial RSS mirror) · Chưa ghi tác giả Mở bản gốc ↗
Nguyên văn nguồn
Chi tiết nguồn
Huấn luyện

How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco

Câu hỏi For this special, we wanted to sit down with someone who has to answer a question a lot of us are asking at work: what should a software engineer actually learn now?

The AI Engineer YouTube channel just crossed one million subscribers . Thank you to everyone who’s watched a talk, sent one to a teammate, or put something they learned into production. For this special, we wanted to sit down with someone who has to answer a question a lot of us are asking at work: what should a software engineer actually learn now? Mihail Eric has two places to test his answer. He teaches The Modern Software Developer at Stanford and leads AI at Monaco, where his colleagues use coding agents to ship software and investigate outages. I asked him to take us through the syllabus, then tell us which parts survived contact with his own engineering team. This is also meant as a preview for all of you joining us in San Francisco for AIE Code next week, as required knowledge for all the conversations with the speakers, sponsors and attendees you’re about to have. If you can see this as a LS subscriber and haven’t got a ticket yet, you can use this code for 30% off (no refunds, sorry) Last year’s homework takes one prompt The first run of CS146S had nine assignments and a capstone. In one exercise, students ran an AI code reviewer on a repo and evaluated its feedback. By the next run, Mihail figured students could one-shot a lot of those exercises. When he asked a room of Stanford sophomores and juniors how they coded, everyone he surveyed was already using an agent. So this time, part of the homework is contributing to an open-source project. The person who maintains the repo reviews the student’s patch. I like what that does to the assignment. You can ask an agent to implement a feature in an afternoon. You still have to find out whether the maintainer wants it, why the code behaves the way it does, and what else your patch might break. That conversation is also how a junior engineer gets to learn from a senior one. There’s an exercise in the episode you can borrow for your next PR. Geoffrey Litt’s explain-diff skill turns a change into a visual walkthrough, then quizzes you about it. Try it on a patch your agent just wrote. Which question do you have to reopen the code to answer? Watch the discussion . What Monaco buys, and what it builds At Monaco, the team evaluated code review products and kept using Devin. Agent workspaces went the other way. A background agent needs somewhere to run the product, reproduce a failure and test the patch. The details depend on how Monaco’s cloud is set up, so the team built its own workspaces. Mihail’s rule of thumb: “You buy to learn, and you build to keep.” One result is Hal. An engineer can ask it to investigate something that broke and open a PR. A salesperson can ask the same system how the product works. In one case the useful answer is a change someone can review. In the other, it’s an explanation you can give a customer. Give both people a wall of debugging output and you’ve only served one of them. The question I kept coming back to was what happens at the other end of the PR. An agent sees the current code. Where does it look to find out why the team rejected this same change six months ago? That takes us from Basis’s Authority Map to Mihail’s prediction that open-source repos will become agent-run software factories. It’s a good reason to send a student to a maintainer today. Someone still has to decide which change belongs in the project. If you’re building with coding agents, come join us at AI Engineer CODE . And if you’ve been teaching this stuff yourself, I’d love to hear which assignment you had to rewrite first. Show notes Stanford CS146S — The ten-week syllabus, assignments and open-source partners. You can also switch to the previous year’s course to compare. Build a coding agent and look inside Claude Code — Mihail’s walkthroughs of the tool loop and the harness around it. Understanding is the new bottleneck — Geoffrey Litt’s approach to learning what changed in an AI-written diff. Monaco’s agent-developer workspaces — The team’s reasoning for building a place where agents can run and test the app. How to Kill the Code Review and the underlying Faros study — Teams with high AI adoption merged 98% more PRs, while their PR review time was 91% longer. Basis’s Authority Map — Where an agent should look for current behavior, earlier decisions and intent. Stripe’s Minions — A production example of giving background coding agents an environment and a path to an engineer-reviewed PR. Boris Cherny’s Steps of AI Adoption — The five stages we pull up on screen near the end of the episode. Mihail on The Thinking Mode — EO Studio’s interview, excerpted at the start of ours. Timestamps 0:00 · One million AI Engineer subscribers and the CODE Summit 0:51 · Mihail Eric and Stanford CS146S 5:26 · What junior engineers need to learn now 10:02 · Open-source maintainers review the students’ code 15:04 · A tour of the ten-week syllabus 24:04 · The loop, tools and harness inside a coding agent 32:18 · Context engineering for a large codebase 38:15 · Skills that make you understand the diff 46:42 · The Monaco.com dispute and the airplane banner 50:23 · What Monaco does for sales teams 59:21 · Inside Monaco’s engineering team 1:01:40 · “You buy to learn, and you build to keep” 1:05:24 · Making repos agent-ready and closing the on-call loop 1:12:37 · When code review becomes the bottleneck 1:20:37 · Security, dependencies and permissions 1:30:13 · Background agents and Stripe’s Minions 1:34:16 · MCP portals and why infrastructure converges 1:41:22 · Building the software factory 1:45:24 · What work is left for developers? 1:52:16 · From personal adoption to team autonomy Transcript Download the complete text below. In the web edition, you can also read the transcript and click any timestamp to jump into the interview. transcript.txt 152.95 KB • PLAIN File Automated transcript, aligned to the final episode. Timestamps open the corresponding moment on YouTube. 0:00 swyx · Intro — Okay, this is a pretty special episode because we're celebrating 1 mi

Nguyên văn nguồn
Chi tiết nguồn
Tác tử

[One Million AI Engineer Special] How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco

Câu hỏi What do you assign when a coding agent can finish last year’s homework in one prompt?

What do you assign when a coding agent can finish last year’s homework in one prompt? In the second run of Stanford’s CS146S, Mihail Eric brought open-source maintainers into the course. Students have to learn an unfamiliar repo, propose a change, and explain the patch to the person who might merge it. We invited Mihail, who teaches The Modern Software Developer at Stanford and leads AI at Monaco, to celebrate one million subscribers on the AI Engineer YouTube channel. He opens up the course syllabus and shows us which lessons survived contact with his own engineering team. Monaco built its agent workspaces, bought its code review tools, and gave an on-call agent access to Datadog and GitHub. We ask how he decides which tools his team needs to own. Watch the full interview and screenshares on YouTube . Show notes Stanford CS146S — The ten-week syllabus and assignments. The Emperor Has No Clothes — Mihail builds a coding agent from scratch. Demystifying Claude Code — Mihail traces a real agent harness. Monaco’s agent workspaces — How the team sets up a workspace for each agent. How to Kill the Code Review — Our essay on what happens when code generation outruns review. Basis’s Authority Map — How agents distinguish current behavior from earlier design decisions. Stripe’s Minions — A look inside Stripe’s background coding agents. AI Engineer CODE Summit — The conference mentioned in the opening. Timestamps 0:00 One million AI Engineer subscribers and the CODE Summit 0:51 Mihail Eric and Stanford CS146S 5:26 What junior engineers need to learn now 10:02 Open-source maintainers review the students’ code 15:04 A tour of the ten-week syllabus 24:04 The loop, tools and harness inside a coding agent 32:18 Context engineering for a large codebase 38:15 Skills that make you understand the diff 46:42 The Monaco.com dispute and the airplane banner 50:23 What Monaco does for sales teams 59:21 Inside Monaco’s engineering team 1:01:40 “You buy to learn, and you build to keep” 1:05:24 Making repos agent-ready and closing the on-call loop 1:12:37 When code review becomes the bottleneck 1:20:37 Security, dependencies and permissions 1:30:13 Background agents and Stripe’s Minions 1:34:16 MCP portals and why infrastructure converges 1:41:22 Building the software factory 1:45:24 What work is left for developers? 1:52:16 From personal adoption to team autonomy Opening excerpt from The Thinking Mode, EO Studio .

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

Will Repos Become Software Factories? — Mihail Eric, Monaco

Câu hỏi What do you assign when students can finish last year’s homework with one prompt?

What do you assign when students can finish last year’s homework with one prompt? For the second run of Stanford’s CS146S, Mihail Eric brought open-source maintainers into the course. Students have to understand an unfamiliar repo, propose a change, and explain it to the person who will review their code. Mihail teaches The Modern Software Developer at Stanford and leads AI at Monaco. We work through his ten-week syllabus, from watching the tool calls inside Claude Code to deciding what a subagent should return to its parent. At Monaco, his team chose to build its own agent workspaces, buy code review tools, and give an on-call agent access to Datadog and GitHub. Mihail explains the reasoning behind those decisions. Mihail thinks open-source repos will become agent-run software factories. If the factory can produce a patch, what should a junior engineer be learning to do? AI Engineer CODE Summit — https://aie.codes AI Engineer on YouTube — https://www.youtube.com/channel/UCLKPca3kwwd-B59HNr-_lvA Chapters 0:00 One million AI Engineer subscribers and the CODE Summit 0:51 Mihail Eric and Stanford CS146S 5:26 What junior engineers need to learn now 10:02 Open-source maintainers review the students’ code 15:04 A tour of the ten-week syllabus 24:04 The loop, tools and harness inside a coding agent 32:18 Context engineering for a large codebase 38:15 Skills that make you understand the diff 46:42 The Monaco.com dispute and the airplane banner 50:23 What Monaco does for sales teams 59:21 Inside Monaco’s engineering team 1:01:40 “You buy to learn, and you build to keep” 1:05:24 Making repos agent-ready and closing the on-call loop 1:12:37 When code review becomes the bottleneck 1:20:37 Security, dependencies and permissions 1:30:13 Background agents and Stripe’s Minions 1:34:16 MCP portals and why infrastructure converges 1:41:22 Building the software factory 1:45:24 What work is left for developers? 1:52:16 From personal adoption to team autonomy Show notes • Stanford CS146S syllabus and assignments — https://themodernsoftware.dev/ • Mihail builds a coding agent from scratch — https://www.mihaileric.com/The-Emperor-Has-No-Clothes/ • Mihail opens up the Claude Code harness — https://www.mihaileric.com/Demystifying-Claude-Code/ • Geoffrey Litt on understanding a diff — https://www.geoffreylitt.com/2026/07/02/understanding-is-the-new-bottleneck.html • Monaco’s agent workspaces — https://www.monaco.com/blog/agent-developer-workspaces • How to Kill the Code Review — https://latent.space/p/reviews-dead/ • Faros report on AI adoption and review time — https://www.faros.ai/blog/ai-software-engineering • Basis’s Authority Map — https://www.getbasis.ai/blogs/how-we-made-our-monorepo-ergonomic-for-agents#canon-vs-not-canon • Stripe’s Minions — https://stripe.dev/blog/minions-stripes-one-shot-end-to-end-coding-agents • Boris Cherny’s Steps of AI Adoption — https://x.com/bcherny/status/2077929379661844559 Opening excerpt from Mihail Eric’s interview on The Thinking Mode, EO Studio — https://www.youtube.com/watch?v=wEsjK3Smovw More Latent Space interviews — https://latent.space

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

v0.34.0rc0: [Transformers backend] Migrate Gemma and StableLM (#66393)

<p>Signed-off-by: Harry Mellor <a href="mailto:19981378+hmellor@users.noreply.github.com">19981378+hmellor@users.noreply.github.com</a><br> Co-authored-by: Claude Opus 5.5 <a href="mailto:noreply@anthropic.com">noreply@anthropic.com</a></p>

Nguyên văn nguồn
Chi tiết nguồn
Đánh giá

Rebuilding GitHub for the Agent Era — Jay Parikh, EVP of CoreAI, Microsoft

Câu hỏi com/badger/home Timestamps 0:00 Why GitHub has to scale for agents 0:26 Introducing Jay Parikh, Microsoft CoreAI 2:35 AI Engineer CODE in San Francisco 3:07 After the GitHub Universe keynote 3:30 One new developer every second 4:52 GitHub as a social network 6:23 Do agents still need pull requests?

GitHub’s next scaling challenge is the agent workload. The platform has 225 million developers and handled 34.3 billion commits over the past year. Jay Parikh, EVP of CoreAI at Microsoft, joins us straight after his GitHub Universe keynote to talk about rebuilding the software-development stack for that scale. We get into Copilot Fusion’s approach to model orchestration and the quality–cost frontier; the harness underneath it; agents that carry a session across phone, desktop and VS Code; and the infrastructure work behind GitHub’s move to Azure. Jay also makes the case for measuring AI spending against completed work and business outcomes. Along the way: why Jay skips pull requests in some of his own repos, Matt Pocock’s repo of Markdown instructions, and the open-source maintenance question behind replacing 2,000 dependencies with 20. Stay for the programmable conference badge at the end. Before the interview: why agent traffic changes GitHub’s job, Jay’s infrastructure background, and AI Engineer CODE — Nov 10–12 in San Francisco. https://ai.engineer/code Show notes • GitHub’s 2026 Octoverse report — The original charts behind the 34.3B commits and agent-first PR figures. https://github.blog/news-insights/octoverse/octoverse-commits-pushes-reach-new-highs-in-2026-as-developers-build-with-agents/ • Matt Pocock’s test-driven development skill — Open the Markdown file in the revision that was current when we recorded. https://github.com/mattpocock/skills/blob/59db609eb1e18aaf59b63284d0e52255aff0c63c/skills/engineering/tdd/SKILL.md • How Copilot Fusion routes a task — GitHub documents when a model drafts alone, when a judge reviews it, and when several models critique it. https://docs.github.com/en/copilot/concepts/models/copilot-fusion • A Copilot session across clients — Watch GitHub’s original demos of a session continuing across devices and VS Code. https://github.blog/changelog/2026-10-28-github-copilot-in-vs-code-universe-releases/ • Push protection for AI-detected secrets — The actual rejected push and what the developer sees in the terminal. https://github.blog/changelog/2026-10-28-push-protection-beyond-predefined-secret-patterns/ • How GitHub migrates Git replicas — GitHub engineers show how they move live copies across data centers and Azure. https://github.blog/engineering/the-bridge-to-agent-scale-git-infrastructure/ • Copilot Automations cost controls — The launch example of per-run, daily, weekly and monthly AI credit limits. https://github.blog/changelog/2026-10-28-new-tools-triggers-and-cost-controls-for-copilot-automations/ • The hackable Universe badge — Source, schematics and games for the badge that swyx holds up. https://github.com/badger/home Timestamps 0:00 Why GitHub has to scale for agents 0:26 Introducing Jay Parikh, Microsoft CoreAI 2:35 AI Engineer CODE in San Francisco 3:07 After the GitHub Universe keynote 3:30 One new developer every second 4:52 GitHub as a social network 6:23 Do agents still need pull requests? 8:03 What counts as code in a repo? 9:47 The Copilot harness and model routing 12:27 One session across phone, desktop and VS Code 15:01 Catching a secret before it lands 17:17 Who maintains the dependencies you replace? 19:10 GitHub in 2027 and the move to Azure 20:51 Token budgets and the value of shipped work 22:36 The programmable Universe badge More from Latent Space: https://www.latent.space

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

v0.40.16-rc0: Add ripple preview support to the MLX runner (#19592)

<p>Add the Ripple model, renderer, parser and import registration.<br> Support per-layer NoPE in DFlash drafts.<br> Fuse MLP activation, value blending and attention gating.<br> Add reference fixtures and model/runner regression tests.</p>

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

Vlad Tenev on Private Investing

You can buy meme coins freely. But private investing is still restricted. Robinhood CEO Vlad Tenev on No Priors: Why accredited investor rules need to change.

Nguyên văn nguồn
Chi tiết nguồn
421–440 / 738
Trang 22 / 37