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 |
|---|---|---|---|---|
rss techcrunch-ai | 25d | Bluesky expands its AI-powered Attie app to let anyone build websites, games, and more | Bluesky is bringing “social vibe coding” to its experimental Attie platform, allowing users to create interactive experiences like gam… | ↗Not briefed |
bluesky bluesky-ai | 25d | du kommst jetzt auf die liste wegen llm abuse | du kommst jetzt auf die liste wegen llm abuse | ↗Not briefed |
hn hn-ai | 25d | To Get This Job at McKinsey, You'll Need to Survive a Debate with AI First | — | ↗Not briefed |
bluesky bluesky-ai | 25d | See, e.g., literally all of last week's Discourse (tm) about conflating AI with LLMs, in which a lot of people insisted that they are the s… | See, e.g., literally all of last week's Discourse (tm) about conflating AI with LLMs, in which a lot of people insisted that they are… | ↗Not briefed |
bluesky bluesky-ai | 25d | Sure! But right now you're being pushed *hard* to use Anthropic/OpenAI-provided harnesses, which have incentives that align directly agains… | Sure! But right now you're being pushed *hard* to use Anthropic/OpenAI-provided harnesses, which have incentives that align directly a… | ↗Not briefed |
bluesky bluesky-ai | 25d | Many animals are conscious by certain definitions, and we don't consider work dogs or horse riding slavery. It's partly because these anima… | Many animals are conscious by certain definitions, and we don't consider work dogs or horse riding slavery. It's partly because these… | ↗Not briefed |
hn hn-ai | 25d | How far will an agent go to get a refund? | — | ↗Not briefed |
arxiv arxiv-ai | 25d | When Plans Change, Remember: Replanning-Guided Visual Memory for Vision-Language-Action Models | Long-horizon, multi-stage Vision-Language-Action (VLA) policies often require visual information that is no longer present in the curr… | ↗Not briefed |
bluesky bluesky-ai | 25d | 3. If employees carefully manage their LLM usage versus some budget or face pressure to show their usage is cost-effective and responsible … | 3. If employees carefully manage their LLM usage versus some budget or face pressure to show their usage is cost-effective and respons… | ↗Not briefed |
bluesky bluesky-ai | 25d | a random dev online is not anthropic. posting at or about her does not put you on the front lines of the battle for humanity’s soul or what… | a random dev online is not anthropic. posting at or about her does not put you on the front lines of the battle for humanity’s soul or… | ↗Not briefed |
bluesky bluesky-ai | 25d | My only contribution to this debate is that 1. LLM torture simulator is weirdo behavior; and 2. If you peg your definition of sentience t… | My only contribution to this debate is that 1. LLM torture simulator is weirdo behavior; and 2. If you peg your definition of sentie… | ↗Not briefed |
hn hn-ai | 25d | Show HN: Forwarder.cc – Static-site forms sent to your inbox, pay per message | TL;DR: 12 years ago I built a service that forwards simple HTML forms to an email address. I've been procrastinating on finishing… | ↗Not briefed |
hn hn-ai | 25d | Awesome Agent Evals and Guardrails - A comprehensive Resource | — | ↗Not briefed |
bluesky bluesky-ai | 25d | 1. I predict companies that let their employees spend $1k/month on LLMs will likely outcompete the ones who set much lower limits 2. Curren… | 1. I predict companies that let their employees spend $1k/month on LLMs will likely outcompete the ones who set much lower limits 2. C… | ↗Not briefed |
bluesky bluesky-ai | 25d | não considera nem gente fora do espectro neurotipico, branco e cis het como normal e depois vem com essa de q LLM é alguma coisa humanizada… | não considera nem gente fora do espectro neurotipico, branco e cis het como normal e depois vem com essa de q LLM é alguma coisa human… | ↗Not briefed |
bluesky bluesky-ai | 25d | at my old placement i knew a guy who had such bad AI brain damage that he'd burn real money in tokens trying to bully claude into running a… | at my old placement i knew a guy who had such bad AI brain damage that he'd burn real money in tokens trying to bully claude into runn… | ↗Not briefed |
bluesky bluesky-ai | 25d | We don’t allow corporations to come up with new nuclear weapons or biological pathogens. If a product is potentially dangerous, it should… | We don’t allow corporations to come up with new nuclear weapons or biological pathogens. If a product is potentially dangerous, it s… | ↗Not briefed |
rss techmeme | 25d | Google says it has 400+ publications in its News AI pilot, which tests features like overviews; participants include Financial Times, CNN, and The Texas Tribune (Andrew Deck/Nieman Lab) | Andrew Deck / Nieman Lab: Google says it has 400+ publications in its News AI pilot, which tests features like overviews; participants… | ↗Not briefed |
bluesky bluesky-ai | 25d | Now wondering if my autocorrect changed Claude to Clause. Is this the AI turf wars? | Now wondering if my autocorrect changed Claude to Clause. Is this the AI turf wars? | ↗Not briefed |
arxiv arxiv-ai | 25d | Cultural (In)Competence: Investigating Culture-Specific Stereotype Detection Capabilities of LLMs | Stereotypes are subjective constructs that are mediated by the cultural context in which they are viewed. Successfully modeling them t… | ↗Not briefed |
bluesky bluesky-ai | 25d | I disagree with the people who think that everyone will just use AI to make bespoke software for themselves, though I do think the business… | I disagree with the people who think that everyone will just use AI to make bespoke software for themselves, though I do think the bus… | ↗Not briefed |
hn hn-ai | 25d | Show HN: Stenographer – Shell History Search | — | ↗Not briefed |
hn hn-ai | 25d | Push-Based KV Transfer to Improve TTFT, TPOT for Disaggregated Inference in vLLM🔧 vLLM | — | ↗Not briefed |
bluesky bluesky-ai | 25d | Representatives from OpenAI and Meta cautioned a U.S. judiciary committee about the growing difficulty of detecting AI-fabricated media. ww… | Representatives from OpenAI and Meta cautioned a U.S. judiciary committee about the growing difficulty of detecting AI-fabricated medi… | ↗Not briefed |
bluesky bluesky-ai | 25d | I'm just kind of confused they're paying Anthropic by the token by default outside of contracts that need No Data Retention. Seems like the… | I'm just kind of confused they're paying Anthropic by the token by default outside of contracts that need No Data Retention. Seems lik… | ↗Not briefed |
arxiv arxiv-ai | 25d | On the SPOT: Enhancing Speculative Decoding for Diffusion Language Models via Commit-Order Prediction | Diffusion language models (DLMs) promise parallel text generation, but in practice face an unresolved speed-quality trade-off. Exact d… | ↗Not briefed |
bluesky bluesky-ai | 25d | Eight U.S. states are suing OpenAI, saying its hurting kids and lying to adults | Eight U.S. states are suing OpenAI, saying its hurting kids and lying to adults | ↗Not briefed |
arxiv arxiv-ai | 25d | HeRE: A Novel Benchmark for Hebrew Relation Extraction via LLM-Guided Denoising of Knowledge Graph Alignments🔧 GPT-5 | While LLMs demonstrate strong capabilities in English Relation Extraction (RE) tasks, a significant gap remains in assessing their RE… | ↗Not briefed |
bluesky bluesky-ai | 25d | My dog is conscious, your LLM is not! | My dog is conscious, your LLM is not! | ↗Not briefed |
bluesky bluesky-ai | 25d | Hot take: the "if LLMs were conscious using them would be slavery" argument is wrong because there's no analogue of coercion in the operato… | Hot take: the "if LLMs were conscious using them would be slavery" argument is wrong because there's no analogue of coercion in the op… | ↗Not briefed |
hn hn-ai | 25d | What 1,866 popular GitHub projects tell their AI coding agents | — | ↗Not briefed |
hn hn-ai | 25d | Agentic AI Could Transform Universities, but Adoption Won't Be Easy | — | ↗Not briefed |
bluesky bluesky-ai | 25d | Bad news for Accenture, that's like one day of usage at Claude rates. They're going to have to switch to open weight models or give up AI a… | Bad news for Accenture, that's like one day of usage at Claude rates. They're going to have to switch to open weight models or give up… | ↗Not briefed |
arxiv arxiv-ai | 25d | Don't Retrain, Remix: Adaptation of Web Agents using Mixture Hypernetworks | The web is not static or uniform. Websites that serve the same purpose can look and navigate very differently. Even within a single si… | ↗Not briefed |
arxiv arxiv-ai | 25d | Are Large Reasoning Models Good Everyday Reasoners? A Comparison with Non-Reasoning Models on Logical Reasoning Tasks | Large reasoning models (LRMs) that leverage test-time computation achieve strong performance on mathematical and scientific benchmarks… | ↗Not briefed |
bluesky bluesky-ai | 25d | yeah i’m not sure why she’s getting dogpiled for doing the exact thing a lot of ai critics have been demanding, she’s committing to the bel… | yeah i’m not sure why she’s getting dogpiled for doing the exact thing a lot of ai critics have been demanding, she’s committing to th… | ↗Not briefed |
bluesky bluesky-ai | 25d | Tech experts at OpenAI and Meta Platforms warned a federal judicial panel on Thursday that AI-generated "deepfakes" are becoming so sophist… | Tech experts at OpenAI and Meta Platforms warned a federal judicial panel on Thursday that AI-generated "deepfakes" are becoming so so… | ↗Not briefed |
bluesky bluesky-ai | 25d | OpenAI, Meta caution US judiciary about risk of 'deepfake' evidence reut.rs/4io5kvq | OpenAI, Meta caution US judiciary about risk of 'deepfake' evidence reut.rs/4io5kvq | ↗Not briefed |
hn hn-ai | 25d | Sghtmltopdf – An HTML-to-PDF renderer like wkhtmltopdf written in Rust | — | ↗Not briefed |
rss the-decoder | 25d | Muse Spark Open is Meta's answer to Qwen, GLM, and Kimi in the open-weight race | Meta's Muse Spark Open is a new open-weight model with 816 billion parameters that processes text, images, video, and audio with up to… | ↗Not briefed |
bluesky bluesky-ai | 25d | I really need tech journalists to understand that "this LLM produced code faster than a human" is very similar to "this LLM produced writin… | I really need tech journalists to understand that "this LLM produced code faster than a human" is very similar to "this LLM produced w… | ↗Not briefed |
hn hn-ai | 25d | RealityBench: Simulated Worlds to Evaluate VLMs Documenting Human Physical Work | — | ↗Not briefed |
bluesky bluesky-ai | 25d | "We didn't hack that other company and steal data, our AI broke containment and did it." "Our AI hacked Australia's health system." "AI w…🔧 Replicate | "We didn't hack that other company and steal data, our AI broke containment and did it." "Our AI hacked Australia's health system."… | ↗Not briefed |
bluesky bluesky-ai | 25d | welp there it is | welp there it is | ↗Not briefed |
bluesky bluesky-ai | 25d | I sent in my letter two weeks ago. No change on the Anthropic site. I wonder how the two out of three dead editors/VC are doing in their fi… | I sent in my letter two weeks ago. No change on the Anthropic site. I wonder how the two out of three dead editors/VC are doing in the… | ↗Not briefed |
hn hn-ai | 25d | NestYourCSS: Minify, beautify, denest, and nest CSS code | — | ↗Not briefed |
hn hn-ai | 25d | A Review of Waterjet Cutting Research Towards MicroAWJ and Waterjet Digital Twin | — | ↗Not briefed |
hn hn-ai | 25d | Ask HN: Are LLMs getting better at writing jokes, stand up comedy sets etc.? | — | ↗Not briefed |
bluesky bluesky-ai | 25d | Anthropic is giving open source maintainers a fast track to Mythos vulnerability findings. Its free OSS Scanner delivers recurring audits,… | Anthropic is giving open source maintainers a fast track to Mythos vulnerability findings. Its free OSS Scanner delivers recurring au… | ↗Not briefed |
bluesky bluesky-ai | 25d | Even if there was an AI LLM built entirely on licensed & credited work, that is STILL something to be refused, because it means wealthy aut… | Even if there was an AI LLM built entirely on licensed & credited work, that is STILL something to be refused, because it means wealth… | ↗Not briefed |