News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
hn hn-ai | 28d | Show HN: Graft – open-source source control for agent-scale monorepos | — | ↗Not briefed |
bluesky bluesky-ai | 28d | So, I picked one of the points that was *very clearly* ripped directly from my own work, and asked him how we came up with it. What sources… | So, I picked one of the points that was *very clearly* ripped directly from my own work, and asked him how we came up with it. What so… | ↗Not briefed |
bluesky bluesky-ai | 28d | So I publicly challenged the guy who published this, and he updated the post to include links to my work as if it's just a handy little ext… | So I publicly challenged the guy who published this, and he updated the post to include links to my work as if it's just a handy littl… | ↗Not briefed |
arxiv arxiv-ai | 28d | MolHop: Scaffold-Hopping Molecular Design with Agentic Reinforcement Learning🔧 GPT-5 | Molecular design is often formulated as an optimization problem over existing molecular spaces. However, molecular optimization and th… | ↗Not briefed |
bluesky bluesky-ai | 28d | there's liability and then the other thing LLM proponents don't seem to understand is: software is about solving problems faced by (sometim… | there's liability and then the other thing LLM proponents don't seem to understand is: software is about solving problems faced by (so… | ↗Not briefed |
rss techmeme | 28d | China's industrial dominance is driven by competitive local governments that act like venture capital firms, making it a continent-sized Silicon Valley (Works in Progress Magazine) | Works in Progress Magazine: China's industrial dominance is driven by competitive local governments that act like venture capital firm… | ↗Not briefed |
arxiv arxiv-ai | 28d | Beyond DFT electronic band structures via dual-space orbital learning | Accurate electronic band structures are essential for predicting material properties, but high-fidelity calculations are computational… | ↗Not briefed |
bluesky bluesky-ai | 28d | "Internetfähige Fernseher, vernetzte Backöfen oder Kühlschränke: Auf Haushaltsgeräten, die mit dem Internet verbunden sind, könnten sich la… | "Internetfähige Fernseher, vernetzte Backöfen oder Kühlschränke: Auf Haushaltsgeräten, die mit dem Internet verbunden sind, könnten si… | ↗Not briefed |
hn hn-ai | 28d | Ask HN: Why do people say "I built it" when we should say "Claude built it"? | It's just something I've been wondering about. Everyone says, “I built it,” but really we should say, “Claude built it for m… | ↗Not briefed |
hn hn-ai | 28d | Grist: Turn the spreadsheets running your business into secure applications | — | ↗Not briefed |
arxiv arxiv-ai | 28d | Not All Criteria Matter: Importance-Aware Criteria Weighting for Rubric-Based LLM Evaluation | Rubric-based evaluation assesses the responses of large language models (LLMs) by aggregating the verification results of explicit cri… | ↗Not briefed |
arxiv arxiv-ai | 28d | MUCO: A Mani-Unit-Structured Manipulation Framework for Compositional Long-Horizon Robot Tasks | Learning and executing long-horizon robot manipulation as complete task trajectories raises data-collection costs, lim?its capability… | ↗Not briefed |
arxiv arxiv-ai | 28d | CARD: Closing the Teacher-Student Loop for Adversarial Robustness Distillation | Adversarial robustness distillation typically treats a robust teacher as a fixed knowledge source and transfers robustness to a compac… | ↗Not briefed |
bluesky bluesky-ai | 28d | The 'dogfooding' culture in the tech industry is so entrenched that this guy used OpenAI's own chatbot to plan his spying on OpenAI. Like,… | The 'dogfooding' culture in the tech industry is so entrenched that this guy used OpenAI's own chatbot to plan his spying on OpenAI.… | ↗Not briefed |
arxiv arxiv-ai | 28d | When Non-Adversarial Noise Hurts Learning: Benchmarking Skill Self-Evolution in Agents | Large language model agents increasingly improve by distilling experience into reusable skills, but errors in that experience can pers… | ↗Not briefed |
arxiv arxiv-ai | 28d | Sparse Transformer: a Controlled Testbed for Mechanistic Interpretability | Mechanistic interpretability aims to decompose neural network computations into human-understandable features, yet progress is limited… | ↗Not briefed |
arxiv arxiv-ai | 28d | Beyond Benchmark Accuracy: A Diagnostic Test for LLM Chemical Prediction | LLMs have achieved strong performance on chemical prediction benchmarks. However, existing evaluation methods typically provide target… | ↗Not briefed |
bluesky bluesky-ai | 28d | The two key events of the day: CPI and Anthropic Q3 numbers. I’d argue the former is micro, while the latter is actually macro. | The two key events of the day: CPI and Anthropic Q3 numbers. I’d argue the former is micro, while the latter is actually macro. | ↗Not briefed |
bluesky bluesky-ai | 28d | There is no "ethical" use of generativeAI. All the commercial platforms have been trained on our stolen books. Did Anthropic write these gu… | There is no "ethical" use of generativeAI. All the commercial platforms have been trained on our stolen books. Did Anthropic write the… | ↗Not briefed |
hn hn-ai | 28d | Show HN: A guided 3D Leonardo da Vinci museum, built in code | Hello HN,<p>I am the founder of a small nonprofit which built Museum of Ages. We opened its first wing last week, a digital Leonardo d… | ↗Not briefed |
rss stratechery | 28d | Apple HomeView, The New Apple TV, The App Story | The Apple Homeview is official, and the question is how big is the addressable market for expensive devices based on subpar AI. | ↗Not briefed |
arxiv arxiv-ai | 28d | Multimodal Medical Image Parsing with Task-Specific Low-Rank Adapters | Adapting a single vision-language model to medical image parsing requires training choices that reflect the demands of each task. We s… | ↗Not briefed |
bluesky bluesky-ai | 28d | OpenAI engineer charged after boardroom bug found wrapped in toilet paper | OpenAI engineer charged after boardroom bug found wrapped in toilet paper | ↗Not briefed |
arxiv arxiv-ai | 28d | How GUI Agents Fail: GraceBench for Execution Stability Diagnosis🔧 GPT-5 | As GUI agents are increasingly deployed for mobile task automation, their reliability depends not only on final task outcomes, but als… | ↗Not briefed |
arxiv arxiv-ai | 28d | Universal Sample-Based Embeddings of Higher-Order Objects | Neural networks operate naturally on finite-dimensional vectors, yet many problems require reasoning about higher-order objects such a… | ↗Not briefed |
bluesky bluesky-ai | 28d | When you bring agents to your hyper malleable text editor, you can plug it to existing tools in your toolbox! Rainbow mode, anyone!? #emacs… | When you bring agents to your hyper malleable text editor, you can plug it to existing tools in your toolbox! Rainbow mode, anyone!? #… | ↗Not briefed |
rss techmeme | 28d | Jensen Huang says Nvidia is pledging $250M in accelerated computing resources over five years to fund UK AI security and agentic safety research (Reuters) | Reuters: Jensen Huang says Nvidia is pledging $250M in accelerated computing resources over five years to fund UK AI security and agen… | ↗Not briefed |
hn hn-ai | 28d | Ask HN: What is the current trajectory of a senior engineer? | Been recently thinking about this more and more: When the need for a human in the design loop ceases?<p>Given that LLMs are RL'd… | ↗Not briefed |
bluesky bluesky-ai | 28d | I thought OpenAI was the sophisticated recording device hidden in our offices? | I thought OpenAI was the sophisticated recording device hidden in our offices? | ↗Not briefed |
bluesky bluesky-ai | 28d | what I *do* think is at serious risk is expensive b2b services that can now be replicated in-house by serious engineers using LLM assistanc… | what I *do* think is at serious risk is expensive b2b services that can now be replicated in-house by serious engineers using LLM assi… | ↗Not briefed |
arxiv arxiv-ai | 28d | DeepPV-MRF: End-to-End Partial-Volume Fraction Estimation from MR Fingerprinting Time Series | Background: Estimating subvoxel tissue fractions from magnetic resonance fingerprinting (MRF) is challenging. Conventional dictionary… | ↗Not briefed |
arxiv arxiv-ai | 28d | Few-Shot Adaptation for Anomalous Sound Detection at Low False-Positive Rate | Reliable fault detection is crucial for industrial machinery in large-scale facilities, where unexpected failures can cause costly dow… | ↗Not briefed |
arxiv arxiv-ai | 28d | On the Limits of Representation-Based Test-Time Adaptation for Multimodal LLMs | Vision-language models suffer significant accuracy drops under common corruptions. While test-time adaptation (TTA) successfully mitig… | ↗Not briefed |
bluesky bluesky-ai | 28d | so: yes, it's now possible for non-coders to create working applications. yes, it's possible for experienced engineers to leverage LLM assi… | so: yes, it's now possible for non-coders to create working applications. yes, it's possible for experienced engineers to leverage LLM… | ↗Not briefed |
bluesky bluesky-ai | 28d | Talks about the importance of accelerating tech adoption and ensuring people don't miss out. Makes analogies with the fear of electricity… | Talks about the importance of accelerating tech adoption and ensuring people don't miss out. Makes analogies with the fear of electr… | ↗Not briefed |
bluesky bluesky-ai | 28d | when I was working in the industry I bought my software, because I was doing runs of books that cost tens of thousands of pounds and I had … | when I was working in the industry I bought my software, because I was doing runs of books that cost tens of thousands of pounds and I… | ↗Not briefed |
bluesky bluesky-ai | 28d | 実はこれも肛門の話をしているようでいて生成AIのベクトル合成の話をしていて、Claudeのロゴというのはその動作機序をよく表しているという話。 RT | 実はこれも肛門の話をしているようでいて生成AIのベクトル合成の話をしていて、Claudeのロゴというのはその動作機序をよく表しているという話。 RT | ↗Not briefed |
arxiv arxiv-ai | 28d | Frustratingly Easy Majority Voting Using Length Conditioning | Majority Voting (MV) is a simple and widely used strategy which improves the reasoning accuracy of Large Language Models (LLMs) by all… | ↗Not briefed |
bluesky bluesky-ai | 28d | I still will avoid the software, since the aforementioned companies have a deeper problem and using an LLM clone only serves to perpetuate … | I still will avoid the software, since the aforementioned companies have a deeper problem and using an LLM clone only serves to perpet… | ↗Not briefed |
bluesky bluesky-ai | 28d | DeepSeek-V4.1-Flash would be a beautiful name for a baby girl | DeepSeek-V4.1-Flash would be a beautiful name for a baby girl | ↗Not briefed |
bluesky bluesky-ai | 28d | A 3D shape that can tile 3D space, leaving no gaps and never forming a repeating pattern, has been discovered using an AI model. The resear… | A 3D shape that can tile 3D space, leaving no gaps and never forming a repeating pattern, has been discovered using an AI model. The r… | ↗Not briefed |
rss techmeme | 28d | CoreWeave says it is targeting insurers' "unlimited" capital, as the AI cloud provider seeks cheaper ways to fund the tens of billions it is spending on chips (Ryan McMorrow/Financial Times) | Ryan McMorrow / Financial Times: CoreWeave says it is targeting insurers' “unlimited” capital, as the AI cloud provider seeks cheaper… | ↗Not briefed |
bluesky bluesky-ai | 28d | "We have to sell solutions, trust and transparency. Not fear. We want to be open and we want to earn trust. But we don't want it to slow u… | "We have to sell solutions, trust and transparency. Not fear. We want to be open and we want to earn trust. But we don't want it to s… | ↗Not briefed |
bluesky bluesky-ai | 28d | I'll say that if people decide that 'your writing has changed in any way over 5 years' is an 'AI tell' we're all going to be in trouble lol… | I'll say that if people decide that 'your writing has changed in any way over 5 years' is an 'AI tell' we're all going to be in troubl… | ↗Not briefed |
bluesky bluesky-ai | 28d | "Trust is everything. It's the soul of OpenAI. We have a model right now that we are not releasing till it's aligned so the folks with mali… | "Trust is everything. It's the soul of OpenAI. We have a model right now that we are not releasing till it's aligned so the folks with… | ↗Not briefed |
bluesky bluesky-ai | 28d | girl that's anthropic 😭😭 | girl that's anthropic 😭😭 | ↗Not briefed |
bluesky bluesky-ai | 28d | Europe keeps naming digital sovereignty as a goal. But how is it put into practice? At today's AI Summit in Dublin, 82% of speaking slots … | Europe keeps naming digital sovereignty as a goal. But how is it put into practice? At today's AI Summit in Dublin, 82% of speaking s… | ↗Not briefed |
hn hn-ai | 28d | Basic Atomic Swap Proof of Concept | — | ↗Not briefed |
bluesky bluesky-ai | 28d | "Ireland's AI adoption is growing fast. You just need to keep doing what you're doing, faster" Sarah Frier of OpenAI at the International A… | "Ireland's AI adoption is growing fast. You just need to keep doing what you're doing, faster" Sarah Frier of OpenAI at the Internatio… | ↗Not briefed |
arxiv arxiv-ai | 28d | Occupancy Modeling via Drifting: One-Step Generation Enables Direct Policy Extraction🔧 Modal | The growing interest in behavior foundation models, combined with recent advancements in generative modeling, has led to an appealing… | ↗Not briefed |