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 | 2mo | If bots take the "internet down", alternatives? | My only thought was: perhaps many, many overly redundant internets that are "Great Firewalled" from one another—serving smal… | ↗Not briefed |
bluesky bluesky-ai | 2mo | This is partly an LLM problem and partly an organizational problem. The LLM makes life much, much easier for everyone who wants to just pre… | This is partly an LLM problem and partly an organizational problem. The LLM makes life much, much easier for everyone who wants to jus… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Autoformalizing Argumentative Material Inferences | Natural language arguments are compelling before they are formally explicit. A premise supports a claim through defeasible warrants, b… | ↗Not briefed |
bluesky bluesky-ai | 2mo | "President Donald Trump met privately with OpenAI chief executive Sam Altman backstage at the Republican midterm convention on Thursday, da… | "President Donald Trump met privately with OpenAI chief executive Sam Altman backstage at the Republican midterm convention on Thursda… | ↗Not briefed |
bluesky bluesky-ai | 2mo | For those who think AI concerns are just a marketing ploy by AI companies, Trump just definitely confirmed their reality when he called the… | For those who think AI concerns are just a marketing ploy by AI companies, Trump just definitely confirmed their reality when he calle… | ↗Not briefed |
rss techmeme | 2mo | Amazon says its India quick delivery service, Amazon Now, crossed $1B in annualized gross revenue, with orders doubling QoQ since its launch in summer 2025 (Sankalp Phartiyal/Bloomberg) | Sankalp Phartiyal / Bloomberg: Amazon says its India quick delivery service, Amazon Now, crossed $1B in annualized gross revenue, with… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Nameless Tokenization: A Lossless Tokenizer-Level Defense Against Control-Token Forgery in Open-Weight LLMs | Open-weight language models publish the strings their chat templates use to mark turns, roles and tool results, which the tokenizer ma… | ↗Not briefed |
bluesky bluesky-ai | 2mo | LET'S TALK TO SOME MORE "STREET EXPERTS"! BBC News is dead. And it seems it wants the same for us. www.bbc.co.uk/news/article... | LET'S TALK TO SOME MORE "STREET EXPERTS"! BBC News is dead. And it seems it wants the same for us. www.bbc.co.uk/news/article... | ↗Not briefed |
hn hn-ai | 2mo | Show HN: GSL – I wanted to describe the architecture, not just draw it🔧 ChatGPT | I built GSL because I got fed up of building architecture diagrams that couldn't quite capture all of the nuance of the real worl… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Es wird an der Stelle logisch, wo dann paar Tage später auf "Hoppla, lasst mal etwas langsamer machen" umgeschwenkt wird. Rein zufällig zu … | Es wird an der Stelle logisch, wo dann paar Tage später auf "Hoppla, lasst mal etwas langsamer machen" umgeschwenkt wird. Rein zufälli… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Structural Negative Transfer in Federated Graph Neural Networks: Diagnosis, Causal Investigation, and the Limits of Divergence-Aware Mitigation🔧 Replicate | Federated learning lets multiple participants train a shared model without pooling raw data, by exchanging locally trained model updat… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Bash sandbox script for Claude and codex CLI | I've been working on and off on this project to have a reasonably easy way to run a docker sandbox on my machine. Recently with c… | ↗Not briefed |
bluesky bluesky-ai | 2mo | For all it's faults, there's one thing LLM has improved. It's made it much easier to tell which people to ignore. I don't even have to watc… | For all it's faults, there's one thing LLM has improved. It's made it much easier to tell which people to ignore. I don't even have to… | ↗Not briefed |
hn hn-ai | 2mo | Composr – a Rust rewrite of PHP's package manager | — | ↗Not briefed |
rss techmeme | 2mo | MediaTek unveils the Dimensity 9600 Pro, its first phone SoC on TSMC's 2nm node, and the 3nm 9600M, with boosted AI performance, as it competes with Qualcomm (Wen-Yee Lee/Reuters) | Wen-Yee Lee / Reuters: MediaTek unveils the Dimensity 9600 Pro, its first phone SoC on TSMC's 2nm node, and the 3nm 9600M, with booste… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 実際のところ「AI()が本物かどうか?」って実は関係なかったりする。 重要なのは「社内と株主を納得させられる稟議をどれだけ回せるか?」ってのが重要で、これって土地でもチューリップでもLLMでもなんでもいいんだよ😩 たまたま「AI」が「これを押さえればMicrosoftやGoo… | 実際のところ「AI()が本物かどうか?」って実は関係なかったりする。 重要なのは「社内と株主を納得させられる稟議をどれだけ回せるか?」ってのが重要で、これって土地でもチューリップでもLLMでもなんでもいいんだよ😩 たまたま「AI」が「これを押さえればMicrosof… | ↗Not briefed |
bluesky bluesky-ai | 2mo | いわゆる世界的というかアメリカの「AIバブル」ってのはまーったくテクノロジーとは無関係で「低金利でガンガン金借りて循環取引でどんどん経済回して膨らませる」っていうスキームだから、イラン戦争から原油の値上がりからの「全世界一斉金利引き上げ」という状況になると「循環取引で経済回すよ… | いわゆる世界的というかアメリカの「AIバブル」ってのはまーったくテクノロジーとは無関係で「低金利でガンガン金借りて循環取引でどんどん経済回して膨らませる」っていうスキームだから、イラン戦争から原油の値上がりからの「全世界一斉金利引き上げ」という状況になると「循環取引で… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Felt like we’d made headway on open source vs proprietary until AI, then DeepSeek open sourced its code and that seemed to freak out the AI… | Felt like we’d made headway on open source vs proprietary until AI, then DeepSeek open sourced its code and that seemed to freak out t… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Splitting the Difference: Interpretable Causal Forests for Treatment Effect Heterogeneity and Bias | In various fields, such as medicine and marketing, accurately predicting individual treatment effects holds significant promise. Howev… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Target-Language Generation in Multilingual Models: Activation Steering and Optimal Control | Ensuring that multilingual language models generate coherent text in a specific target language is a major issue in multilingual langu… | ↗Not briefed |
arxiv arxiv-ai | 2mo | HUMAID-NER: A Disaster Tweet Dataset for Joint Named Entity Recognition and Event Classification via Uncertainty-Weighted Multitask Learning | Rapid extraction of structured information from social media is important for humanitarian response, yet existing disaster tweet resou… | ↗Not briefed |
bluesky bluesky-ai | 2mo | He later served as White House chief of staff. On Saturday, Emanuel seized on the warnings from Anthropic CEO Dario Amodei. His essay, Eman… | He later served as White House chief of staff. On Saturday, Emanuel seized on the warnings from Anthropic CEO Dario Amodei. His essay,… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Affect-Prototype Guided Fusion for Open-Vocabulary Incomplete Multi-modal Emotion Recognition🔧 Modal | Open-vocabulary multimodal emotion recognition (OV-MER) aims to generate open natural-language emotion labels from multimodal affectiv… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Pourquoi vous reprenez la com de anthropic ? | Pourquoi vous reprenez la com de anthropic ? | ↗Not briefed |
hn hn-ai | 2mo | Motivation of Vieta's Transformation (2015) | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | ちゃんと「学習データ(LLM)は、万引きしたデータで作られています。超巨大検索デマ合成エンジンですよ、盗んだデータを使って切り貼りしているんです、ときに個人情報も抜き出せます」 って教えてもらいましたか?ん?😊 分かったかな、小学生たちー?😉 #生成AI法規制スタンディン… | ちゃんと「学習データ(LLM)は、万引きしたデータで作られています。超巨大検索デマ合成エンジンですよ、盗んだデータを使って切り貼りしているんです、ときに個人情報も抜き出せます」 って教えてもらいましたか?ん?😊 分かったかな、小学生たちー?😉 #生成AI法規制ス… | ↗Not briefed |
rss the-decoder | 2mo | Apple brings a fully revamped Siri built on Google's Gemini, but not to the EU | Apple is shipping its rebuilt "Siri AI" after years of delay, built on Google's Gemini models and running partly on the device, partly… | ↗Not briefed |
hn hn-ai | 2mo | YC batches got twice as likely to repeat an earlier YC company, from 18% to 47% | — | ↗Not briefed |
hn hn-ai | 2mo | Should AI friendliness be a priority for your projects? | Recently I came up with an idea that if MCP server and AI focused readme is reasonable for your project than it must be done.<p>Take f… | ↗Not briefed |
rss techmeme | 2mo | Microsoft rolls out emergency fix for critical issues caused by its September Patch Tuesday update, which addressed ~1,000 vulnerabilities but introduced bugs (Tom Warren/The Verge) | Tom Warren / The Verge: Microsoft rolls out emergency fix for critical issues caused by its September Patch Tuesday update, which addr… | ↗Not briefed |
hn hn-ai | 2mo | ZTE B860H: Getting Armbian into It | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | AntennaFlow: A Generative Flow Model for Offset Correction in Phaseless Antenna Testing | Near-field to far-field transformation is central to large-aperture antenna testing, yet two coupled challenges remain: costly phase a… | ↗Not briefed |
arxiv arxiv-ai | 2mo | AeroLat: Channel-Aware Latent Space Semantic Communication for Decentralized UAV Swarms | Communication in latent space offers an intriguing alternative to symbolic messages for decentralized autonomous Unmanned Aerial Vehic… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 😵💫 Donald označil obavy z AI za hoax a nacpal k tomu i změnu klimatu nebo Ukrajinu. Řekl bych, že Dario Amodei z Anthropicu, který veře… | 😵💫 Donald označil obavy z AI za hoax a nacpal k tomu i změnu klimatu nebo Ukrajinu. Řekl bych, že Dario Amodei z Anthropicu, který… | ↗Not briefed |
hn hn-ai | 2mo | Measuring Gauss-Seidel loop-carried dependency and fixing it via loop unrolling | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Trump says AI safety fears a 'hoax' as he rejects calls for greater safeguards. www.bbc.com/news/article... | Trump says AI safety fears a 'hoax' as he rejects calls for greater safeguards. www.bbc.com/news/article... | ↗Not briefed |
bluesky bluesky-ai | 2mo | KI-Unternehmen wie #OpenAI, #Google und #Perplexity schließen Verträge mit Medien, um journalistische Inhalte zu nutzen. Über 70 Prozent de… | KI-Unternehmen wie #OpenAI, #Google und #Perplexity schließen Verträge mit Medien, um journalistische Inhalte zu nutzen. Über 70 Proze… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The latest weekly release does not include the AJAX Spider and DOM XSS add-ons. You can install them via the Marketplace as usual, but we r… | The latest weekly release does not include the AJAX Spider and DOM XSS add-ons. You can install them via the Marketplace as usual, but… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Beyond Token-Local Imitation: Reward-Compatible Temporal Credit Assignment for On-Policy Distillation | On-policy distillation (OPD) has emerged as an effective approach for large language model post-training, yet existing objectives face… | ↗Not briefed |
arxiv arxiv-ai | 2mo | RepoAtlas: Guiding Coding Agents via Evolving Multimodal Repository Views | Large language model (LLM)-powered coding agents have made rapid progress in automating software engineering tasks, yet repository-lev… | ↗Not briefed |
hn hn-ai | 2mo | WebGPU Engine Bench: Three.js vs. PlayCanvas vs. Babylon.js | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | MedPCFM-TED: One-Step Point Cloud Flow Matching for Implant Generation via Teacher-Guided Endpoint Distillation | Cranial implant generation is an important task in medical imaging. Recent point cloud based generative methods, particularly flow mat… | ↗Not briefed |
rss techmeme | 2mo | China tightens overseas travel controls for citizens via exit bans and fines in a sweeping new law, as Beijing moves to secure state secrets and advanced tech (Financial Times) | Financial Times: China tightens overseas travel controls for citizens via exit bans and fines in a sweeping new law, as Beijing moves… | ↗Not briefed |
arxiv arxiv-ai | 2mo | When Confidence Signals Disagree: Local and Global Confidence in Autoregressive Language Models🔧 Modal | Modern predictive systems expose multiple quantities that are commonly interpreted as measures of confidence. However, these quantitie… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Causal Discovery via Transformed Low-Rank Quantile Surfaces | We propose Low-Rank Quantile Surfaces (LRQS), a bivariate causal model in which, in the causal direction, an unknown monotone transfor… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Repurposing Deep Limit Order Book Forecasting for Scenario-Conditioned Market Impact Modeling | Deep Limit Order Book forecasting models capture nonlinear market dynamics, but their ability to quantify the effects of counterfactua… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Verbalizing Subliminal Learning Effects Using Text Optimization | Subliminal learning is a phenomenon in which a distillation dataset transmits traits from the teacher model that are not legibly encod… | ↗Not briefed |
arxiv arxiv-ai | 2mo | HyCoSeq: Contextual Hyperbolic Representation Learning for Genomic Sequences | Hyperbolic geometry provides a natural inductive bias for genomic representation learning, but existing hyperbolic genomic models prim… | ↗Not briefed |
rss techmeme | 2mo | Google DeepMind AI Safety and Alignment researcher Bilal Chughtai publicly resigns, saying "I earnestly believe that AI has the potential to kill us all" (Debby Wu/Bloomberg) | Debby Wu / Bloomberg: Google DeepMind AI Safety and Alignment researcher Bilal Chughtai publicly resigns, saying “I earnestly believe… | ↗Not briefed |
rss mit-tech-review-ai | 2mo | What must happen for AI’s trillion-dollar gamble to pay off | When Jessica Wachter, a finance professor at the University of Pennsylvania’s Wharton School, wanted to assess AI’s impact on the econ… | ↗Not briefed |