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

Kỹ thuật

Launching an opt-in vulnerability-finding service for open-source software

We’re making available OSS Scanner, an opt-in vulnerability scanner for the open-source ecosystem that’s informed by our experience using Claude to find vulnerabilities during Project Glasswing.

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

The missing map of the sky

Brice Ménard, an astrophysicist at Johns Hopkins University and a researcher at Anthropic, explains how he worked with Claude Science to produce the first complete map of the sky in UV light.

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

OpenAI Dots Took Over Our Company

Subscribe to Every (free for 14 days): https://every.to/youtube-every-offer?utm_source=podcast&utm_medium=audio&utm_campaign=companyagentepisode On this week's episode of The Every Podcast, Dan Shipper sits down with Willie Williams, Every's head of platform and one of the driving forces behind the Every Agent, which launched this week. They discuss how Every went from everyone running their own OpenClaw to relying on one shared company agent—and where they think personal agents belong in everyday life. Links to resources mentioned in the episode: • Willie Williams on X: https://x.com/bigwilliestyle • The Every Agent: https://every.to/agent?utm_source=podcast&utm_campaign=podcast&utm_content=1006companyagent • OpenClaw: https://openclaw.ai • Tend, Dan's open-source feed app: https://every.to/tend?utm_source=podcast&utm_campaign=podcast&utm_content=1006companyagent • Every’s Vibe Check on Dots: https://every.to/vibe-check/vibe-check-dots-always-on-agents-in-chatgpt?utm_source=podcast&utm_campaign=podcast&utm_content=1006companyagent • Dot from OpenAI: https://openai.com/index/introducing-dots/ 0:00 Introduction 5:58 How OpenClaw spread through Every 11:13 Why everyone's personal agents eventually died 14:51 The difference between building for builders and building for operators 16:20 Why work will run on company agents while personal agents will stay personal 19:56 How engineering management has changed in the age of AI 21:43 Using an AI-powered feed to see everything happening across a product launch 23:53 Why Willie's token usage fell off a cliff 29:05 The benefit of having separate AI coaches for each part of your life 31:45 How Every is building its agent to surface information unprompted

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

Introducing Playground: Create and play custom games

<img src="https://storage.googleapis.com/gweb-uniblog-publish-prod/images/THUMBNAIL_BLOG.max-600x600.format-webp.webp">Playground is a new experimental gaming platform that lets you create, play, and share custom games.

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

Open-Source AI & Open Models Reading List

Hey all! I’ve been prepping for some public-audience and policy-facing writing on open models, so I figured I would share my research materials. There’s lots of wonderful stuff in here. This is my list of the best writing on open models in the last few years. If someone decides they want to get up to speed on the area, reading this will be a comprehensive overview of the state of affairs. Please comment pieces to consider adding below, and I’ll update this over time. List last updated: 26 Oct. 2026 Share Foundation What open models are, why people release them, how they relate to business strategy, and what the risks are. On open source AI strategy, a walkthrough of how open-source software has been used by businesses and early signs of what that means for AI — From Open Source Software to Open Source Strategy , Bill Gurley (May 2026). One of the clearest articulations is Mark Zuckerberg’s comments around Llama 3’s release as to why Meta releases open models — Open Source AI is the Path Forward , Mark Zuckerberg (Jul. 2024) Why you should view open models on a gradient, rather than binary open/closed, based on factors such as licenses, cost of running the model, data access, etc. — The Gradient of Generative AI Release: Methods and Considerations , Irene Solaiman (Feb. 2023). The role open models will play in the economy of the future, as a complement to strong closed models. Why open models will be used to create custom agentic workflows in enterprises across the world – What comes next with open models , Nathan Lambert / Interconnects (Mar. 2026) A position on how open models will capture value by providing a complementary tool to large swaths of the existing economy, drawing on the history of IP and current debates on open vs. closed models (e.g. distillation) — Some Simple Economics of Open versus Closed AI , Christian Catalini (Aug. 2026) Why open models will constantly be behind closed models in performance — Open models in perpetual catch-up , Nathan Lambert / Interconnects (Feb. 2026) Where adoption differs for open and closed models — Open and closed models are on different exponentials , Nathan Lambert / Interconnects (Jun. 2026) A clear articulation on how to balance releasing powerful open-weight models while taking safety seriously — A Safe Path to Open Weights , Thinking Machines Lab (Jul. 2026). Early paper on marginal risks that showed text-focused LLMs very marginally increased documented potential risks of models — On the Societal Impact of Open Foundation Models , Sayash Kapoor, Rishi Bommasani et al. (Feb. 2024). Closed models safety guardrails are regularly bypassed causing a plethora of real AI-risks before hypothetical risks of open weight models have emerged — The Myth of unsafe Open Source AI , Florian Brand (Jun. 2026). The mass reduction in open data, which is a crucial factor that has hampered truly open AI research — Consent in Crisis: The Rapid Decline of the AI Data Commons , Shayne Longpre et al. (Jul. 2024). Recent examples on how strong Chinese models impact the AI ecosystem — Kimi K3: The open-weights escalation , Nathan Lambert / Interconnects (Jul. 2026) / GLM-5.2 is the step change for open agents , Nathan Lambert / Interconnects (Jun. 2026). A summary of the story of open models in 2025: Nathan Lambert on China’s AI Ecosystem and the Open Model Gap | The Curve 2025 , Golden Gate Institute for AI (Nov. 2025). A lecture surveying all of the above topics: Open-source history, licenses, open model history, adoption numbers, open risks, and balancing the benefits of openness. Nathan Lambert — YouTube (Oct. 2026) [ Optional ] Latest data on open model adoption: A general summary on US vs. China model adoption — The ATOM Report (Apr. 2026), The latest data on model downloads, derivatives, and research adoption by region — Interconnects Adoption Dashboard , and The most important models to know about in the ecosystem — Interconnects Artifacts Hub Interconnects AI is a reader-supported publication. Consider becoming a subscriber. US-China Competition Who is leading in open models, how this has changed over time, how China maintains its leading position, and relevant history. Why the U.S. needs to invest in open models for fundamental R&D / innovation in the face of growing competition from China – The ATOM Project , Nathan Lambert (Aug. 2025) The lens as to why open models help spur research innovation and beneficial outcomes for AI — Why I build open language models , Nathan Lambert / Interconnects (Oct. 2024) Why open models foster education, innovation and competition, three core American values — Banning Open Source AI Would Be A Mistake , Nathan Lambert & Kevin Xu (Jun. 2026) Why the recent “vibe regulation” / vague federal oversight mechanisms set us up for a clash and-or ban of frontier open models in the near future — 6 months to live for open models , Nathan Lambert / Interconnects (Jul. 2026) [ Optional ] Fully open language model technical reports to illustrate the start of the art in understanding: Pythia (EleutherAI, 2023), Olmo (2024), Olmo 2 (2024), Olmo 3 (2025) Chinese open-source history leading up to AI — Chinese Open Source: A Definitive History , Kevin Xu (Mar. 2026). China’s structural advantages in open-source — China’s Structural Advantage in Open Source AI , Kevin Xu (Jun. 2025). How Chinese labs themselves discuss building models, and how the Chinese industry differs from the U.S. — Notes from inside China’s AI labs , Nathan Lambert / Interconnects (May 2026). Why Chinese labs are so good at keeping up with American competition (e.g. American open weight labs struggle to compete with Chinese labs on fair performance comparisons) — GLM-5.3: How Chinese labs keep stride with the frontier , Nathan Lambert / Interconnects (Aug. 2026). Prominent uses of Chinese models by Western companies have prompted meaningful regulatory attention ( more discussion ) Lawmakers have probed the following companies over using Chinese models: DoorDash ( C

  • The mass reduction in open data, which is a crucial factor that has hampered truly open AI research — Consent in Crisis: The Rapid Decline of the AI Data Commons , Shayne Longpre et al.
Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

Introducing Claude Haiku 5.5

Claude Haiku 5.5 is our fastest, most capable small model. Built for high-volume work like summarization, subagents, and browser use.

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

Release v5.19.0

<h1>Release v5.19.0</h1> <h2>New Model additions</h2> <h3>EmbeddingGemma2</h3> <a target="_blank" rel="noopener noreferrer" href="https://private-user-images.githubusercontent.com/73884904/666833754-84734e74-163d-4d12-b166-ffcf6749d563.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.EBwQw9OqDo69b6vmhMPUEeU22Y9Va_6R5JuJuvgWh7w"><img width="2000" height="1700" alt="image" src="https://private-user-images.githubusercontent.com/73884904/666833754-84734e74-163d-4d12-b166-ffcf6749d563.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3OTUwNjk2MTgsIm5iZiI6MTc5NTA2OTMxOCwicGF0aCI6Ii83Mzg4NDkwNC82NjY4MzM3NTQtODQ3MzRlNzQtMTYzZC00ZDEyLWIxNjYtZmZjZjY3NDlkNTYzLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNjExMTklMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjYxMTE5VDA2MjE1OFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTU0ODUwOWM0MDg0MjczYzI3YWMyMjg1OGM0NjI4YjhjM2Y3OTZjNDc4ZjhlM2UyNDk1ZGYyZTg0NDgxZTZkNzAmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0JnJlc3BvbnNlLWNvbnRlbnQtdHlwZT1pbWFnZSUyRnBuZyJ9.EBwQw9OqDo69b6vmhMPUEeU22Y9Va_6R5JuJuvgWh7w" content-type-secured-asset="image/png" style="max-width: 100%; height: auto; max-height: 1700px;"></a> <p>EmbeddingGemma 2 is a multimodal embedding model from Google built on the Gemma 4 architecture. It encodes text, images, audio, and video, individually or combined in one input, into a shared 768-dimensional vector space for cross-modal retrieval, semantic similarity, clustering, and classification. It uses Matryoshka Representation Learning, so embeddings can be truncated to 512, 256, or 128 dimensions. It also offers configurable visual and video token budgets, and unused vision or audio towers can be disabled at load time to save memory.</p> <p><strong>Links:</strong> <a href="https://huggingface.co/docs/transformers/main/en/model_doc/embedding_gemma2" rel="nofollow">Documentation</a></p> <ul> <li>Smthn smthn (<a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5731273262" data-permission-text="Title is private" data-url="https://github.com/huggingface/transformers/issues/49364" data-hovercard-type="pull_request" data-hovercard-url="/huggingface/transformers/pull/49364/hovercard" href="https://github.com/huggingface/transformers/pull/49364">#49364</a>) by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/vasqu/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/vasqu">@vasqu</a> in <a href="https://github.com/huggingface/transformers/pull/49364" data-hovercard-type="pull_request" data-hovercard-url="/huggingface/transformers/pull/49364/hovercard">#49364</a></li> </ul> <h2>Breaking changes</h2> <p>All MoE models whose routers compute logits now return them when <code>output_router_logits=True</code>, following the Qwen3-MoE pattern (a <code>router_logits</code> recorder on the base model, <code>MoeModelOutputWithPast</code> from the backbone, and a MoE causal LM output from the head), so code that relied on the previous outputs or their absence should read the router logits from these output classes.</p> <ul> <li>🚨 Return router logits from every MoE model that computes them (<a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5495317954" data-permission-text="Title is private" data-url="https://github.com/huggingface/transformers/issues/48920" data-hovercard-type="pull_request" data-hovercard-url="/huggingface/transformers/pull/48920/hovercard" href="https://github.com/huggingface/transformers/pull/48920">#48920</a>) by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/qgallouedec/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/qgallouedec">@qgallouedec</a></li> </ul> <p><code>Owlv2ForObjectDetection.embed_image_query</code> now selects the query box with the highest objectness score, as in the original OWLv2 notebook, instead of the OWL-ViT heuristic, so image-guided query embeddings and detections may differ from earlier releases.</p> <ul> <li>🚨 Select the OWLv2 image query by objectness (<a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5640150994" data-permission-text="Title is private" data-url="https://github.com/huggingface/transformers/issues/49200" data-hovercard-type="pull_request" data-hovercard-url="/huggingface/transformers/pull/49200/hovercard" href="https://github.com/huggingface/transformers/pull/49200">#49200</a>) by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/qgallouedec/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/qgallouedec">@qgallouedec</a></li> </ul> <p>The <code>"paged|"</code> prefix for SDPA and flash attention implementations is deprecated, so users should set the regular attention implementation (e.g. <code>sdpa</code> or <code>flash_attention_2</code>) for continuous batching instead of <code>paged|sdpa</code> or <code>paged|flash_attention_2</code>.</p> <ul> <li>🚨 Attention 🚨 Deprecate "paged|" prefix for SDPA and flash (<a class="issue-link js

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

The Cyber Risk Discourse is Broken

Câu hỏi The problem with this is not the technical research they did, which is largely reasonable, but the failure to engage on more cross cutting questions like: What happens if we ban open models due to cyber risks or why do Chinese companies deem these models safe to release?

I’ve had one too many discussions on the cyber risks of open models — where someone assumes that either banning open models happens in a silo (and bad actors will somehow actually be stopped) or that China is a threat and doesn’t care about AI safety — that I feel we’re going to end up making policy decisions that both limit American AI competitiveness and increases long-term cyber risk. I feel like we’re locked in on a path of lose-lose situation, unless people embrace nuance, trade-offs, and scary realities. Much has been written about the views of various players in this debate, which I summarize as follows: Open weight models pose an untenable risk to society’s functionality , occupied by frontier lab leadership and the national security community in the U.S. Western voices saying open weight models are necessary for defense and banning them will make the world less safe , occupied by AI risk moderates to different extremes. I include myself in this bucket, along with Hugging Face’s perspective after the OpenAI incident, and others like Joshua Saxe . Chinese companies continuing to release open-weight models with strong cyber capabilities , who are making decisions based on the risk assessments of their society and government. This list is also organized by volume or clarity of message in the Western AI ecosystem. The risk-focused discussions have been disproportionately visible, including pieces of work which I view as very detrimental to potential of engaging across assumed party lines on this issue. There’s also very little substantive engagement trying to understand how Chinese companies assess risk — rather, mostly arguments that border on ad hominem attacks like “Chinese labs don’t care about safety.” Interconnects AI is a reader-supported publication. Consider becoming a subscriber. The delusions of the anti open-weight alliance The most recent publication from the “open-weights are dangerous” crowd was the report by Anthropic on the risks of GLM-5.3 as an offensive cyber tool. The problem with this is not the technical research they did, which is largely reasonable, but the failure to engage on more cross cutting questions like: What happens if we ban open models due to cyber risks or why do Chinese companies deem these models safe to release? These questions, which I’ll come back to, are the ones we need to answer to understand the ecosystem perspective on current AI risks. All policy actions happen in a multi-piece puzzle, where the shape of risk can be changed, but it is rare that you get a universally better outcome. They are all trade-offs. This sort of solipsistic writing by Anthropic fits very closely to another type of discourse I’m hearing about regularly that is determining the fate of open models and AI broadly — classified briefings. I’ve had many discussions with people who say something like “Look, I’m pro open models, but if you’re seeing what I’m seeing, it’s an onslaught out there due to [XYZ Chinese open-weight model] and we need to act.” This perspective flies in the face of public information, where to date closed models have been documented as the cause of most existing cyber attacks. As someone who tries to make grounded predictions, the evidence of numerous attacks from OpenAI models is the only data I have on the shape of cyber risks (or, you can broaden the scope to include all of FelonyBench ). There are a few possible explanations, and it’ll take years to get to the true answer: Maybe open model weights and closed model APIs (with some safe guards) are both far closer to being easy to mis-use, rather than API models being closer to safe. The trope "Open Dangerous, Closed Safe" may be closer to "Open Unsafe, Closed Unsafe." Maybe there are far fewer bad actors who are willing to use AI models for “loud” cyber attacks that target crucial infrastructure in the US or powerful and underprepared institutions. Closed models have stronger capabilities and stronger safeguards (in theory), but the stronger capabilities part could matter more in net harm if the safeguards on both are porous. I have a lot more to learn about how the cybersecurity ecosystem works, so I won’t support this full-heartedly as one of my current recommendations, but I think it’s at least a reasonable argument that open weight models as a diffusion tool for cyber defense could be the best tool we have to prevent harm over the next few years. There are many domains, e.g. sensitive government agencies, where open-weight models are the only tools that can be deployed in the near-term on air-gapped networks. Lots of this situation seems like the messy reality being hard to pin down, where the logical limits are clearer. I may agree with the simple reality that closed models are far safer in the limit, as a technical tool with more layers of protection, but they may actually be causing more harm in the near term because of how accessible model APIs are. Still, much of the impact of the tools comes down to how they’re used and controlled. A duopoly on such a crucial tool may be the defining, risky part. This is the debate we’re having, and over time, as intelligence diffuses, many things may change. These together make me come to the conclusion that if you think the latest open-weight models need to be banned to slow the diffusion of cyber risks, you probably also need to make public-facing APIs for the frontier closed-models illegal. The current stack of safeguards on closed models is stronger than open-weight models, but far from perfect. It is very likely that cyber capabilities of closed models increase much faster than guardrail performance, and a world where open models are banned while closed models continue to progress would be increasing the offense-defense cyber gap. It will take meaningful time to build mitigations that allow defenders to use strong AI models on their private infrastructure if we remove access to strong open-weight models. Share Explaining China’s AI risk posture

Nguyên văn nguồn
Chi tiết nguồn
Huấn luyện

Applied Compute CEO on the Limits of RL, the New AI Hyperscaler & Why Post-Training Wins Inference

Câu hỏi " 0:00 Intro 1:42 The Case for Owning Your Intelligence 4:35 OpenAI x Baseten and the Multi-Model Future 7:19 Can Post-Training Beat Frontier Models?

Yash Patil, a former OpenAI researcher who worked on Codex and now runs Applied Compute, argues that post-training is the most overlooked part of the AI stack and the key to owning your intelligence. He explains when companies should optimize their harness and context first, and when it's worth updating the weights: the largest inference workloads, plus tasks that change over time or depend on judgment. That's the heart of why he believes post-training wins inference. The biggest workloads are both the most valuable to train and the most valuable to serve, and how a model is trained shapes how it should be served. Getting production serving running is the easier half; echoing Dylan Patel, he says GPUs plus vLLM plus OpenRouter "kind of" gets you there. Yash is candid about the limits of RL. It's a hill-climbing machine whose hardest part is defining the hill, which makes a company's private evals as worth protecting as its employees. It generalizes far less than pre-training. And continual learning remains blocked by the unsolved problem of extremely data-efficient training from sparse rewards. He lays out his bet on a new AI hyperscaler, built on GPUs the way AWS, GCP, and Azure were built on CPUs. He pictures it as an inverted pyramid with training at the bottom, where few can compete, and inference, routing, and harness stacked above. The conversation also covers whether every firm needs its own model, why Jevons paradox has made cost matter more than he expected, how models reward hack "like water," and why the US needs American open-weight models even though "blanket bans are not the answer." 0:00 Intro 1:42 The Case for Owning Your Intelligence 4:35 OpenAI x Baseten and the Multi-Model Future 7:19 Can Post-Training Beat Frontier Models? 9:03 What Data Stays Out of Distribution 11:51 Data Efficiency and Continual Learning 14:31 When Updating Model Weights Makes Sense 16:01 Online Training in Practice 18:16 Lowering the Cost of Post-Training 19:48 Does Every Company Need Its Own Model? 24:06 RL in Non-Verifiable Domains 26:59 How Applied Compute Picks Customers 28:46 Building an Inference Business 34:43 The Case for a New AI Hyperscaler 38:21 Building RL Environments 41:33 Quickfire Jacob is an AI investor at Redpoint Ventures. He's led Redpoint's investments in companies like Abridge, Physical Intelligence & Legora. Follow Jacob on Twitter: https://x.com/jacobeffron On Unsupervised Learning we probe the sharpest minds in AI in search for the truth about what's real today, what will be real in the future and what it all means for businesses and the world. If you're a builder, researcher or investor navigating the AI world, this podcast will help you deconstruct and understand the most important breakthroughs and see a clearer picture of reality. Subscribe to this show to stay up to date on our latest episodes.

Unsupervised Learning · Unsupervised Learning: With Jacob Effron Mở bản gốc ↗
Nguyên văn nguồn
Chi tiết nguồn
Huấn luyện

Google's AI Infrastructure Chief, Amin Vahdat, on the Physics & Economics of Frontier AI

At 100,000-accelerator scale, something fails multiple times an hour, which is why Google's Chief Technologist for AI Infrastructure Amin Vahdat thinks FLOPS is a vanity metric. The metric that matters is what he calls goodput: the useful work a workload actually delivers through real failures. Amin walks through the calculus that split the TPU line into 8i and 8t for the first time, why the TPU's core primitives haven't changed since v1, and how Google and DeepMind co-design in the same rooms, intercepting chip architectures mid-flight before tape-out. He explains why long-horizon agents are sending demand for CPUs and storage through the roof alongside accelerators, how optical circuit switches reroute light to a spare rack in milliseconds, and why Google would rather wait on a utility than build its own gigawatt. We also cover orbital data centers and the multi-megawatt rack of 2036. Hosted by Sonya Huang, Sequoia Capital 0:00 – Introduction 1:47 – What makes a data center an AI data center 5:30 – Goodput, not FLOPS: holding yourself accountable when something fails every hour 11:52 – Doubling token capacity every six months, and where the gains actually come from 15:32 – The TPU bet: from a contrarian call in 2013 to splitting 8i and 8t 23:30 – The case for and against co-design 26:11 – Shoulder to shoulder with DeepMind: intercepting chips mid-flight 34:16 – Long-horizon agents change the shape of the data center 37:50 – Optical circuit switching and the state of networking 42:35 – Power is the binding constraint: utilities, gigawatts, and sizing a data center 49:23 – Training vs. serving clusters, seven-year-old TPUs, and open standards 58:29 – Orbital data centers and the supercomputer of 2036

Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

One resignation turned the embers of AI fear into a wildfire

Câu hỏi Why did this quitting announcement reach so far?

As AI became more powerful, it was inevitable that a different, growing group would start to take AI safety more seriously – what we did not know ahead of time, is which set of views they latched onto. We have seen that some of the most extreme views of risk, i.e. moderate probabilities of mass extinction, were the ones that reached the masses. A lot in the AI world is about to change due to this. How did we get here? Why did this quitting announcement reach so far? In many ways, the rest of the world’s views around AI in the past was a dampening factor. You can think about this like the damp ground around a fire. Many people were striking matches for years about AI risk – they’d smolder in their community and largely burn out, going unnoticed. As the stakes of AI have risen this year, from the OpenAI-HuggingFace incident and breakthroughs like the Navier-Stokes result (also from OpenAI), the ground has dried out and the latent energy around the AI discourse has increased. More people not in the industry have thought, “huh, maybe I should care about this AI thing.” The ambient temperature and stakes have been obviously rising. Then, some basic factors of human nature apply, with the most crucial being that fear sells. Fear is the simplest story, the one people cannot look away from. Jacob Coxon was the one who stumbled into this new powder keg, totally unaware of what was going to come. What looked like a fairly innocuous event – another AI researcher quitting citing safety risks – landed into a very different environment and it caught like wildfire. The discussion of existential risk, mass extinction, and the trajectory of AI has traveled further than even the most seasoned AI commentariat would ever predict. Share There are a set of facts we need to get clear, which paint the picture of the situation. The key Tweets to reference are from Jacob Coxon , the resignation thread, and Evan Hubinger , the source of the >10% extinction risk figure . There are plenty of AI risks which are likely to cause harm, even if estimating annihilation is useless. It is important to weigh these with respect to the benefits. The entire discourse around existential risk is on very poor footing. At least Evan was clear in his post, with “kill all humans,” but a major problem in the AI Safety discourse is that people talk about existential risks, when they mean very different things (much like how AGI is a vaguely meaningless term ). I put the probability of complete extinction as being so low it isn’t worth discussing, but the probabilities of AI caused disasters – e.g. cyber attacks on critical infrastructure or bio-risks – as being worth debating. Throwing this whole discussion out because there are not these disasters yet is a harmful reaction. Jacob Coxon is acting genuinely and with good intentions. The outpouring of support from more well-established AI researchers who know of him and his intentions of resignation is useful. Many factions of AI turned to scapegoating him individually, based on account metadata, personal factors, etc. These are not useful. Many frontier lab employees genuinely have similar views to him. I’m not sure it’s a majority, but there is a substantial group. Many frontier lab employees, especially at Anthropic, are out of touch and this will impact their forecasting and/or descriptions of current AI events. I say this without blaming individuals, but it’s a common agreement among my friends not at OpenAI/Anthropic (Ant especially) that people at the labs operate with a religious energy. It’s very common to go through very out of touch interactions with them. I do not blame most of the individuals who get distorted views being part of these companies, but the interactions are wild and spill over into a lot of wack discussions in the AI media ecosystem. Living in this environment that normalizes such out of touch behavior will inevitably distort any human’s understanding of technical progress. Interconnects AI is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. This was not a mass political campaign, but rather an opportunistic media coordination. For context, the Wall Street Journal had an exclusive story that Jacob coordinated before posting. I suspect that Jacob shared his plan of quitting in groupchats with AI safety advocacy groups ahead of time, e.g. the morning of posting, asking for amplification. This is normal practice, and could have included some prominent politicians . From there, I think it’s more likely that other politicians are bandwagoning on a rising issue. When you combine this with other factors, like Daniel Kokotajlo’s appearance on Joe Rogan coming out the same day – it definitely looks like a very well-executed, coordinated media campaign. This doesn’t mean it’s a conspiracy or a regulatory capture tactic within Democratic political structures. The determining factor seems to be that no one – including Jacob and those posting about X-risk today – knew that it would go so viral. We do not have proof that RSI causes the risks these researchers forecast . The general argument for RSI follows as: The current pace of progress is very high, the current progress is heavily dependent on AI tools, the current AI tools are superhuman in some domains (e.g. math) – so, all together, AI is going to work more on itself and become superhuman in all relevant areas over time to autonomy and intellect. This view dramatically undersells human bottlenecks in building models and allocating resources at organizations, and draws conclusions on future AI capabilities more broadly. I called my alternative view to this, Lossy self-improvement . AI has always been very jagged, and we are making models which are superhuman goal-seekers at math and software engineering, but they have massive limitations on intuitions, creativity, and other types of reasoning that humans are strong at. With AI agents assisting researc

  • As the stakes of AI have risen this year, from the OpenAI-HuggingFace incident and breakthroughs like the Navier-Stokes result (also from OpenAI), the ground has dried out and the latent energy around the AI discourse has increased.
  • I called my alternative view to this, Lossy self-improvement .
Nguyên văn nguồn
Chi tiết nguồn
Tác tử

Remote control for local agents

You can now see and reply to the local agents running on your computer from the Cursor iOS app. Get started Remote Control is on by default for everyone except Enterprise organizations. Download the Cursor iOS app and sign in. Computers on your account appear in the app automatically. Tap your computer in the app and approve your pairing request in the Cursor desktop app. Your local agents appear in a list. Tap one to see what it's doing or send it a message. <figure><img src="https://ptht05hbb1ssoooe.public.blob.vercel-storage.com/assets/changelog/remote-control-cl-pezewO3SZRrIQxMNaTzdpsP9TD1ChS.png" loading="lazy" alt="Remote Control for local agents in the Cursor iOS app" /><figcaption>Remote Control for local agents in the Cursor iOS app</figcaption></figure> Agents keep running on your computer Remote control doesn't move your agents anywhere. It keeps running on your computer, and the app connects to it. Your computer needs to stay on and online for remote control to work. Keep your computer awake To stop your computer from sleeping while you're away, turn on "Keep this computer awake" under Remote Control in Cursor's desktop settings. Your computer needs to be plugged in with the lid open. For Enterprise teams Remote control is on by default, except for Enterprise organizations. Enterprise admins can turn it on in Org settings > Security and identity > Remote control. Remote control doesn't require cloud agents to work. Remote control is available today in the Cursor iOS app .

Nguyên văn nguồn
Chi tiết nguồn
Kỹ thuật

2.3.1

<table> <thead> <tr> <th>Package</th> <th>Version</th> </tr> </thead> <tbody> <tr> <td><code>@modelcontextprotocol/client</code></td> <td>2.3.1</td> </tr> <tr> <td><code>@modelcontextprotocol/server</code></td> <td>2.3.1</td> </tr> <tr> <td><code>@modelcontextprotocol/core</code></td> <td>2.3.1</td> </tr> <tr> <td><code>@modelcontextprotocol/server-legacy</code></td> <td>2.3.1</td> </tr> <tr> <td><code>@modelcontextprotocol/codemod</code></td> <td>2.3.1</td> </tr> <tr> <td><code>@modelcontextprotocol/node</code>, <code>express</code>, <code>hono</code>, <code>fastify</code></td> <td>unchanged</td> </tr> </tbody> </table> <h2>Changes</h2> <ul> <li><code>requireBearerAuth</code> in <code>@modelcontextprotocol/server-legacy</code> takes the optional <code>expectedResource</code> that <code>@modelcontextprotocol/server</code> 2.3.0 added: it accepts only tokens issued for this server (the token's audience). Off unless you set it. (<a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5710576774" data-permission-text="Title is private" data-url="https://github.com/modelcontextprotocol/typescript-sdk/issues/2952" data-hovercard-type="pull_request" data-hovercard-url="/modelcontextprotocol/typescript-sdk/pull/2952/hovercard" href="https://github.com/modelcontextprotocol/typescript-sdk/pull/2952">#2952</a>)</li> <li>The npm pages of <code>@modelcontextprotocol/server</code> and <code>@modelcontextprotocol/client</code> now open with one note that links the documentation, the migration guide and the issue form. (<a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5711246129" data-permission-text="Title is private" data-url="https://github.com/modelcontextprotocol/typescript-sdk/issues/2955" data-hovercard-type="pull_request" data-hovercard-url="/modelcontextprotocol/typescript-sdk/pull/2955/hovercard" href="https://github.com/modelcontextprotocol/typescript-sdk/pull/2955">#2955</a>)</li> </ul>

Nguyên văn nguồn
Chi tiết nguồn
  • 2.3.1 modelcontextprotocol/typescript-sdk releases · 5 thg 10, 2026
Kỹ thuật

1.32.1

<h2>What's Changed</h2> <ul> <li>[v1.x] docs: state the principles in CLAUDE.md by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/claude/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/claude">@claude</a>[bot] in <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5681372162" data-permission-text="Title is private" data-url="https://github.com/modelcontextprotocol/typescript-sdk/issues/2939" data-hovercard-type="pull_request" data-hovercard-url="/modelcontextprotocol/typescript-sdk/pull/2939/hovercard" href="https://github.com/modelcontextprotocol/typescript-sdk/pull/2939">#2939</a></li> <li>[v1.x] docs: point the README at v2 and say what v1.x supports by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/claude/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/claude">@claude</a>[bot] in <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5682278405" data-permission-text="Title is private" data-url="https://github.com/modelcontextprotocol/typescript-sdk/issues/2942" data-hovercard-type="pull_request" data-hovercard-url="/modelcontextprotocol/typescript-sdk/pull/2942/hovercard" href="https://github.com/modelcontextprotocol/typescript-sdk/pull/2942">#2942</a></li> <li>chore: bump version to 1.32.1 by <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/users/claude/hovercard" data-octo-click="hovercard-link-click" data-octo-dimensions="link_type:self" href="https://github.com/claude">@claude</a>[bot] in <a class="issue-link js-issue-link" data-error-text="Failed to load title" data-id="5711243793" data-permission-text="Title is private" data-url="https://github.com/modelcontextprotocol/typescript-sdk/issues/2954" data-hovercard-type="pull_request" data-hovercard-url="/modelcontextprotocol/typescript-sdk/pull/2954/hovercard" href="https://github.com/modelcontextprotocol/typescript-sdk/pull/2954">#2954</a></li> </ul> <p><strong>Full Changelog</strong>: <a class="commit-link" href="https://github.com/modelcontextprotocol/typescript-sdk/compare/1.32.0...1.32.1"><tt>1.32.0...1.32.1</tt></a></p>

Nguyên văn nguồn
Chi tiết nguồn
  • 1.32.1 modelcontextprotocol/typescript-sdk releases · 5 thg 10, 2026
Đánh giá

GPT-6 Astra, Looped Transformers, and Hidden Reasoning

Câu hỏi But what, exactly, has it improved, and how?

A lot has happened in the last few weeks. I am sure that OpenAI’s GPT-6 Astra is top of mind for everyone right now. In particular, thoughts on its performance, the looped transformer/recurrent depth aspects, and rumors that Astra is “hiding” its reasoning trace (i.e., chain of thought). So, in this article, I want to start with some brief impressions of Astra and some thoughts on where all this is headed. Then, I will discuss, in detail, what “looped transformers” are, and how (or rather, if) this relates to hiding chains of thought. Lastly, after covering the basics of the looped transformer, I wanted to highlight some new insights from recent research papers on the topic. PS: I also have a shorter video version here that I recorded before writing this more extended article: 1. GPT-6 Astra impressions First things first. Before getting into the architecture rumors and related research literature, let me briefly summarize some GPT-6 Astra observations and tidbits. Last week, OpenAI’s new GPT-6 Astra was released with a big fanfare. I used it over the past couple of days, and it’s an exceptionally good model, likely the best I’ve used as of this writing. But what, exactly, has it improved, and how? 1.1 Astra benchmarks Astra is the best model I’ve used so far, and it’s disproportionately good at 3D rendering and animation tasks (relative to other models). With that, I mean that while it leapfrogs its GPT-5.6 predecessor in practically all categories (writing, math, coding, and more), it especially does so when it comes to graphical demos. We can see this also reflected in the benchmarks. For instance, GPT-6 Astra is really good at math and coding, as shown below. Figure 1: Selection of three popular coding benchmarks and one challenging math benchmark. More benchmarks are shared on the Astra release blog: https://openai.com/index/gpt-6-astra/ One of the highlights (not shown in the figure) is that Astra also achieves 99.9% on the ARC-AGI-3 benchmark (GPT-5.6 Sol only 7.8%), which measures a mix of solving logic puzzles and generalization. However, the math, coding, and computer use benchmarks are more interesting because they are closer to real-world use. (However, there was then also an article by OpenAI describing two further changes to the harness that also tripled GPT 5.6 Sol’s performance from to 13.3% to 38.3%.) Coming back to the Artificial Analysis Coding Agent Index v1.4 (lower right in the previous figure), which blends several agentic coding tasks, GPT-6 Astra is clearly at the frontier, but it doesn’t pull ahead by leaps and bounds. This can also be seen in the general Artificial Analysis Intelligence Index shown below, which blends different types of tasks, not just coding tasks. Figure 2: Artificial Analysis Intelligence Index via https://artificialanalysis.ai/#intelligence Now, the big advantage of Artificial Analysis benchmarks is that they are independent and thus may be a bit more trustworthy than self-evaluated benchmarks by model developers. The harness setup depends on the benchmark . For example, GDPval-AA and AA-Briefcase use their open-source, minimal Stirrup harness across the different LLMs they compare. In the Intelligence Index v4.2 shown above, Terminal-Bench v2.1 uses Terminus 2, and τ³-Banking uses the τ-Bench harness. The separate Coding Agent Index also compares different coding-agent harnesses. For evaluations that use a shared harness, this makes it more of an apples-to-apples comparison. At the same time, during model training, models are typically developed with one primary harness in mind (and fine-tuned less on other harnesses). Plus, the primary harness is often developed to suit and amplify a model’s strengths. So, some of the agentic evaluations might underestimate how well Astra performs in its primary harness. How much this affects its Intelligence Index score would need to be tested by comparing Astra across harnesses on the same tasks. As a side note, as a colleague recently suggested to me (as also recommended by the Claude Code lead), it’s maybe not a bad idea to delete (/archive) some of your existing AGENTS.md contents and SKILL.md files, as newer LLMs have become more efficient at understanding the prompt and solving the problem at hand. The extra hand-holding could unnecessarily constrain newer models and lead to worse solutions. Of course, I am not suggesting never using SKILL.md files again, but for some workflows, because they can improve efficiency upon reuse, since the model doesn’t have to rediscover them. But what I am suggesting is that some workflows don’t need describing, and “old” descriptions may no longer be ideal, and the LLM may be able to come up with better solutions. So, it’s perhaps time to update or regenerate said instruction files. 1.2 Computer use capabilities GPT-6 Astra seems to be exceptionally strong in image and rendering tasks. When these tasks involve interacting with graphical user interfaces, they also demonstrate computer-use capabilities, meaning the model operates software on your local computer through the Codex/ChatGPT app. Computer use is where the model really shines compared to others, and anything graphic-related also makes for interesting and intuitive demos on social media platforms. There are tons of examples of impressive demos out there, from modeling rendering New York City in blender to virtual open house tours . To pick one example, below is a comparison where I had GPT-6 Astra Medium and High redraw a picture of me in a browser version of MS Paint using the mouse on my computer (not Extra High and Max, because I didn’t want to waste all my tokens :)). This highlights not only the model’s artistic capabilities but, more importantly, its ability to use tools on one’s computer (in this case, Paint; you can see the model using the interface via the mouse cursor). This is not the first model that, inside a harness, is capable of general computer use. For example, I successfully used GPT model

  • md files again, but for some workflows, because they can improve efficiency upon reuse, since the model doesn’t have to rediscover them.
Nguyên văn nguồn
Chi tiết nguồn
Kiến trúc

When will average people feel AI’s impact?

Câu hỏi How will a common citizen come to credit OpenAI or Anthropic for saving their life, if they went to their family doctor who told them about a new miracle cure?

Housekeeping: Paid subscribers to Interconnects now get a permanent 40% discount on my book when purchasing at Manning.com. Access the code at the Interconnects perks page . Many AI optimists tend to compare what is happening in this AI boom to the industrial revolution, or to other periods of rapid technological advancement and diffusion into society. These comparisons fit on the scale of technological change, but miss a crucial factor in how most people are exposed to that change. The problem facing AI is that most people have no super tangible new goods thanks to it and society has more inertia resisting change than in previous eras. Share I’m writing this coming back online from a few weeks off for my wedding in New England. In this time it would have been very easy to not think about AI at all. The touch-points that average people have to AI products today are fringe, marginally beneficial, or even just very confusing to them (e.g. many people have heard about and brought up the OpenAI-HuggingFace incident, but don’t know what to make of it). People on the positive side think of AI as a way to make fun images, enhanced Google Search, etc. These are very small benefits. On the negative side is an association with addictive social media algorithms, friends of friends addicted to AI chatbots, and a plethora of takes on data centers. AI is still a rounding error in everyday life Core aspects of everyday life — family, food, transportation, and entertainment — have few direct impacts yet. It’s a remarkable breath of fresh air to pop out of the bubble and realize how little what is happening really matters today . Being obsessed with AI is a choice that a very few people have yet opted into. For example, the only thing I used AI for in this time was search and creative work (making the pretty seating chart for my wedding guests to find their table). In industrial revolutions past, average people got absolutely life changing outcomes. The First Industrial Revolution in the late 18th century gave access to cheaper clothing, cooking ware, reading material, and a shift to new livelihoods. The Second Industrial Revolution in the late 19th century introduced household machines (e.g. sewing machines), preserved food, indoor plumbing, photography, better light sources, bicycles, and further benefits of manufactured goods and electrification. The list is remarkable — most of these we still use regularly today — and very physical. While even the most optimistic versions of AI will usher in new scientific discoveries, advanced therapeutics for rare diseases, and potentially even sustained economic abundance, these benefits have the risk of being too indirect. How will a common citizen come to credit OpenAI or Anthropic for saving their life, if they went to their family doctor who told them about a new miracle cure? What percentage of Americans will care about OpenAI solving the Navier-Stokes Millennium Prize Problem ? It feels very likely in 50 years that the average American’s day to day life looks very similar. Their home, appliances, relationships, and vehicles will be similar (of course, self-driving will continue to diffuse, but that has been developing on a very independent trajectory from the innovations of LLMs). In this time, AI will get a lot of credit. 50 years is a remarkable length of time with how fast everything is changing today in this, AI-focused narrow slice of the world. The most important part of what is happening early in the AI revolution, is building foundational infrastructure, and a general process, which will compound over decades. A major mathematical breakthrough today will look astonishingly minor in scope relative to the advancements that come later in the compounding journey. It is hard to predict what it looks like for every technology you use daily to get faster compounding improvements due to AI. Interconnects AI is a reader-supported publication. Consider becoming a subscriber. Breaking social stasis Much of the narrative around AI is trying to push people to care, due to this long-term reality of progress, at least as a subconscious motive. This will take a long, long time to get right, and AI’s buildout faces immediate political problems due to this imbalance. Today’s AI is primarily a tool to serve the elite. For knowledge work, which is roughly half of the U.S. economy, AI is as fundamental as electricity (or quickly will be so, with rapid improvements to agents in the next 18 months). It’s highly destabilizing to have such a transformative, productive tool only bring half of society along. It’s not hard for many people to pick up on this — the technology economy booms while life stays otherwise stagnant. In writing this, I learned of Engels’ pause , which is “the period from 1790 to 1840, when British working-class wages stagnated and per-capita gross domestic product expanded rapidly during a technological upheaval.” 1 If we — the leaders of the AI industry — think this is the closest analogue to what comes next for AI, those not benefiting are right to push back. AI is the greatest tool ever for scaling technology companies and starting new online-native small businesses. I don’t even expect the tech industry to grow in headcount and nurture its workers through an era of massive success — I agree with that headcount would likely shrink while knowledge work output explodes . These sectors were already the most successful in the American economic system, so the brand of AI will be tarnished as not being a collective good. I worry that this instinctive reaction will kneecap AI’s development, sending it down a path that looks closer to the cautionary tale of American nuclear power. Part of the challenge is the speed and relentlessness of expectations in society. The AI industry has millions of eyes on it, and won’t get much patience to wait and bring innovations later. If given 100 years to diffuse into society, its impacts will certainly become much more obvi

  • It is hard to predict what it looks like for every technology you use daily to get faster compounding improvements due to AI.
  • economy, AI is as fundamental as electricity (or quickly will be so, with rapid improvements to agents in the next 18 months).
Nguyên văn nguồn
Chi tiết nguồn
Tác tử

Together Link: open models in the harness you already use. Start with one command today.

Together Link brings frontier open models like GLM 5.3 and Kimi K3 into the coding agent your team already uses, cutting model spend by over 50%.

Nguyên văn nguồn
Chi tiết nguồn
641–660 / 738
Trang 33 / 37