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 |
|---|---|---|---|---|
bluesky bluesky-ai | 11d | my rule at work for my direct reports is "AI written code is fine if you can tell me exactly what everything does". i knew it was fucking o… | my rule at work for my direct reports is "AI written code is fine if you can tell me exactly what everything does". i knew it was fuck… | ↗Not briefed |
bluesky bluesky-ai | 11d | one of the great parables of our time: Tesla has spent the better part of a decade trying to make camera-based "AI" auto wipers a thing, in… | one of the great parables of our time: Tesla has spent the better part of a decade trying to make camera-based "AI" auto wipers a thin… | ↗Not briefed |
arxiv arxiv-ai | 11d | Federated Learning for Cross-Regional Epidemic Forecasting with Asynchronous Outbreaks | Objective: Epidemic outbreaks are often asynchronous across regions, yet collaborative forecasting is hampered by privacy laws and het… | ↗Not briefed |
bluesky bluesky-ai | 11d | on the ramp index, total spend is up 10% sept-oct a lot of it seems to basically be market share from openAI. total usage was also bonkers … | on the ramp index, total spend is up 10% sept-oct a lot of it seems to basically be market share from openAI. total usage was also bon… | ↗Not briefed |
hn hn-ai | 11d | BasalFit – Recipes/meal plans and food log, usable from Claude-ChatGPT via MCP🔧 ChatGPT | — | ↗Not briefed |
bluesky bluesky-ai | 11d | It's a shame they hit a point that they had to ask an LLM for help fixing a serious performance issue after Unity themselves couldn't provi… | It's a shame they hit a point that they had to ask an LLM for help fixing a serious performance issue after Unity themselves couldn't… | ↗Not briefed |
hn hn-ai | 11d | Tell HN: Agent that control everyscreen beat human in clash royale | Hi HN!!<p>I am developer of AutoCua ( automatic computer use agent).<p>Its a computer use agent , that can automate. iphone, andriod,… | ↗Not briefed |
hn hn-ai | 11d | The Multimodal Knowledge Agent Benchmark | — | ↗Not briefed |
bluesky bluesky-ai | 11d | If people dislike the game for having LLM generated code, there is nothing wrong with that. I personally hate it myself. I just think inva… | If people dislike the game for having LLM generated code, there is nothing wrong with that. I personally hate it myself. I just think… | ↗Not briefed |
hn hn-ai | 11d | Gemini 4 Carbon is rolling out today | — | ↗Not briefed |
bluesky bluesky-ai | 11d | I don’t support House House’s decision to use an LLM in this case, but I get why they did it. And I think there’s some irony in having thi… | I don’t support House House’s decision to use an LLM in this case, but I get why they did it. And I think there’s some irony in havin… | ↗Not briefed |
arxiv arxiv-ai | 11d | Achieving Robust Channel Estimation Neural Networks on Doppler effects by Designed Training Data | Channel estimation is crucial in wireless communication systems. Many existing works train and test their models using the same or sim… | ↗Not briefed |
bluesky bluesky-ai | 11d | People are treating it like the entire game is AI created slop, which I don't think it is. Are people wrong for being upset with undisclose… | People are treating it like the entire game is AI created slop, which I don't think it is. Are people wrong for being upset with undis… | ↗Not briefed |
bluesky bluesky-ai | 11d | The core of a university, in the relevant sense, is the library. And in a very literal sense, LLM-based artificial intelligence could not e… | The core of a university, in the relevant sense, is the library. And in a very literal sense, LLM-based artificial intelligence could… | ↗Not briefed |
arxiv arxiv-ai | 11d | Physics-Constrained Autoencoder for Interpretable Health Representation Learning in Turbofan Engines | Health Indices (HIs) summarize degradation from operational data and support diagnostic and prognostic tasks in Prognostics and Health… | ↗Not briefed |
arxiv arxiv-ai | 11d | Chain of Physics: Supporting Fault Diagnosis with Engineering-Inspired Large Language Model Reasoning | Fault diagnosis in complex systems requires linking observed changes in physical quantities to the failure mechanisms that produce the… | ↗Not briefed |
hn hn-ai | 11d | The State of Mechanistic Interpretability for AI Safety | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Recurrent Modular Structure in Spiking Neural Network Reinforcement Learning | Biological neural circuits are both modular and recurrent, with dense communities sparsely linked to one another. Spiking neural netwo… | ↗Not briefed |
bluesky bluesky-ai | 11d | "we have to use LLMs to find information because all other avenues have been destroyed" yes, *by the LLM companies*, this is one of the rea… | "we have to use LLMs to find information because all other avenues have been destroyed" yes, *by the LLM companies*, this is one of th… | ↗Not briefed |
hn hn-ai | 11d | Gemini 4 Carbon is now available | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Lilo: A Multi-Agent Conversational Sandbox for Children's Digital Communication | Children's peer conflicts increasingly play out in group chats, where tone and timing are stripped away, and a hurtful message is easi… | ↗Not briefed |
hn hn-ai | 11d | Show HN: A Swift package that makes any macOS app translatable | As you can see on my "The low-tech guys" [0] website, I create and maintain a bit too many apps as a single person. I never… | ↗Not briefed |
bluesky bluesky-ai | 11d | The bigger question is, why should I give up RAM, good video cards, a functioning open web and advertising model that enables worldwide col… | The bigger question is, why should I give up RAM, good video cards, a functioning open web and advertising model that enables worldwid… | ↗Not briefed |
bluesky bluesky-ai | 11d | Counterpoint: looking for bugs in your game is billable hours you're paying Anthropic for in tokens instead of your employees in wages. | Counterpoint: looking for bugs in your game is billable hours you're paying Anthropic for in tokens instead of your employees in wages. | ↗Not briefed |
bluesky bluesky-ai | 11d | The important thing to keep in mind about the Big Walk LLM usage is you *cannot*, and I mean under any circumstances, rip on ANYONE for hav… | The important thing to keep in mind about the Big Walk LLM usage is you *cannot*, and I mean under any circumstances, rip on ANYONE fo… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Ilo Musi 8 – Handheld Game Console Emulating CHIP-8 Platform | Hello HN, I built ilo musi 8. It's a handheld device that emulates CHIP-8 gaming platform. It took me almost a year to build it.<… | ↗Not briefed |
arxiv arxiv-ai | 11d | A Cross-Scenario Survey and Analysis of Self-Evolving Multimodal Agents | Recent work shows that multimodal agents increasingly self-improve without recurring human annotation. However, each self-improvement… | ↗Not briefed |
bluesky bluesky-ai | 11d | Sources: Goldman Sachs expects Anthropic revenue of ~$52B in 2026 and $600B+ by 2030, operating profit in 2029, and positive free cash flow… | Sources: Goldman Sachs expects Anthropic revenue of ~$52B in 2026 and $600B+ by 2030, operating profit in 2029, and positive free cash… | ↗Not briefed |
rss deepmind-blog | 11d | Gemini 4 Carbon is now available | — | ↗Not briefed |
rss techmeme | 11d | Sources: Goldman Sachs expects Anthropic revenue of ~$52B in 2026 and $600B+ by 2030, operating profit in 2029, and positive free cash flow in 2030 (Financial Times) | Financial Times: Sources: Goldman Sachs expects Anthropic revenue of ~$52B in 2026 and $600B+ by 2030, operating profit in 2029, and p… | ↗Not briefed |
bluesky bluesky-ai | 11d | AI is evil and run by fascists but *my* use is ok and its not my fault im reliant on it. I hope they fix the world but keep the part I like… | AI is evil and run by fascists but *my* use is ok and its not my fault im reliant on it. I hope they fix the world but keep the part I… | ↗Not briefed |
arxiv arxiv-ai | 11d | Semantic Chains: From Reachable Facts to Justified Answers | How can the facts that justify an answer be found from a question and presented as a sequence of connected statements? This paper mode… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Ghost Runner – Race your past self on Apple Watch | Making running somewhat of a routine has the side effect of repeating the same routes very often. As someone raised playing games like… | ↗Not briefed |
bluesky bluesky-ai | 11d | AFAICT, weirdly enough, big walk was basically _not_ made with LLM assistance, except for them asking a chatbot a few questions ONCE that … | AFAICT, weirdly enough, big walk was basically _not_ made with LLM assistance, except for them asking a chatbot a few questions ONCE… | ↗Not briefed |
bluesky bluesky-ai | 11d | genuinely what do you say about this | genuinely what do you say about this | ↗Not briefed |
bluesky bluesky-ai | 11d | An LLM will always give you a statistically average response and jump you straight to the end, or near the end. To me thinking through the … | An LLM will always give you a statistically average response and jump you straight to the end, or near the end. To me thinking through… | ↗Not briefed |
rss theverge-ai | 11d | Google releases Gemini 4 Carbon with agent ‘swarm’ feature | Google is rolling out its latest AI model, Gemini 4 Carbon, to the general public after releasing it to a select group of "cybersecuri… | ↗Not briefed |
bluesky bluesky-ai | 11d | I’ll just say that the use case house house is describing for big walk is one of the LLM uses I can pretty easily get behind for any sort o… | I’ll just say that the use case house house is describing for big walk is one of the LLM uses I can pretty easily get behind for any s… | ↗Not briefed |
bluesky bluesky-ai | 11d | Everyone seems to think AI can write philosophy papers, but when I ask Claude Sonnet 5 a philosophy question, it cites AI-generated fake en… | Everyone seems to think AI can write philosophy papers, but when I ask Claude Sonnet 5 a philosophy question, it cites AI-generated fa… | ↗Not briefed |
bluesky bluesky-ai | 11d | me: well we can always look at the upsides amiright anthropic: actually *no* | me: well we can always look at the upsides amiright anthropic: actually *no* | ↗Not briefed |
bluesky bluesky-ai | 11d | Anthropic didn't have any choice but to launch this deal, despite the headwinds, and they will maintain a game face until and unless the ma… | Anthropic didn't have any choice but to launch this deal, despite the headwinds, and they will maintain a game face until and unless t… | ↗Not briefed |
hn hn-ai | 11d | RapidFort Secures the AI Stack from GPU Runtime to Kubernetes | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Personalized Affective Perception via Continual Learning in Human-Agent Negotiation | Self-interested intelligent agents that negotiate with humans should interpret nonverbal expressions and adapt their behavior accordin… | ↗Not briefed |
arxiv arxiv-ai | 11d | SinkState: On-the-Fly SVD of Recurrent States for Efficient Hybrid LLM Serving | Hybrid models combine softmax attention with linear attention, which compresses history into fixed-size recurrent states to balance mo… | ↗Not briefed |
arxiv arxiv-ai | 11d | An AI Assistant for Automating Anti-Corruption Expertise of Draft Legislation Based on the TRACE Framework | We present an AI assistant for automating anti-corruption expertise of draft legislation based on the TRACE architecture, designed to… | ↗Not briefed |
bluesky bluesky-ai | 11d | Will be a real test of open weight models. If in six weeks there is a model that can decompile as good as Opus can then the moat arround An… | Will be a real test of open weight models. If in six weeks there is a model that can decompile as good as Opus can then the moat arrou… | ↗Not briefed |
bluesky bluesky-ai | 11d | Google broadens its AI hardware portfolio with two new specialized chips | Google broadens its AI hardware portfolio with two new specialized chips | ↗Not briefed |
bluesky bluesky-ai | 11d | I'm a big AI skeptic, but I've gotta be honest: this situation where any admitted use of AI — even as a glorified search engine to solve a … | I'm a big AI skeptic, but I've gotta be honest: this situation where any admitted use of AI — even as a glorified search engine to sol… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Ziperu is rubyzip with a Rust core, same API but faster and safer | Annoyed by slow zip handling in Ruby apps in gems like roo and caxlsx, they use rubyzip underneath which is pure Ruby and does a lot o… | ↗Not briefed |
rss techcrunch-ai | 11d | Google broadens its AI hardware portfolio with two new specialized chips | Google's new tensor processing units (TPUs) broaden its chip portfolio to more use cases: confidential model training and more memory… | ↗Not briefed |