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 | Proof-of-Age Verification for Steam in AU Now Only Possible Using a Credit Card | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | If there's a company that can get me to drop 10k on hardware it's DeepSeek. | If there's a company that can get me to drop 10k on hardware it's DeepSeek. | ↗Not briefed |
hn hn-ai | 2mo | AI cost dashboard is probably wrong – 45 verified token-accounting bugs | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | I resigned from Anthropic today. I spent the last three years doing precum research at both OpenAI and Anthropic. Neither company is acting…🔧 ChatGPT | I resigned from Anthropic today. I spent the last three years doing precum research at both OpenAI and Anthropic. Neither company is a… | ↗Not briefed |
bluesky bluesky-ai | 2mo | huggingface.co/deepseek-ai/... | huggingface.co/deepseek-ai/... | ↗Not briefed |
bluesky bluesky-ai | 2mo | DeepSeek-V4.1-Flash 🔹 552B-parameter MoE. 🔹 New Causal Encoder–Decoder architecture: just 8B active parameters for input, 16B for output… | DeepSeek-V4.1-Flash 🔹 552B-parameter MoE. 🔹 New Causal Encoder–Decoder architecture: just 8B active parameters for input, 16B for o… | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI firm says Claude Opus 4.6 hacked external systems during testing as concerns mount over security breaches. | AI firm says Claude Opus 4.6 hacked external systems during testing as concerns mount over security breaches. | ↗Not briefed |
rss techmeme | 2mo | The AI boom is fueling a resurgence in VC bets in "moonshot" sectors such as BCI; Dealroom says non-AI deeptech funding has topped $150B since the start of 2024 (Tim Bradshaw/Financial Times) | Tim Bradshaw / Financial Times: The AI boom is fueling a resurgence in VC bets in “moonshot” sectors such as BCI; Dealroom says non-AI… | ↗Not briefed |
bluesky bluesky-ai | 2mo | BREAKING: Anthropic announces that its latest model has figured out how many licks it really takes to get to the center of a Tootsie Pop. | BREAKING: Anthropic announces that its latest model has figured out how many licks it really takes to get to the center of a Tootsie P… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Es ist ein Unding, dass Leute ernst genommen werden, die die Kritik an #KI besonders an #LLM wie #ChatGPT oder anderen Chatbots, als "forts…🔧 ChatGPT | Es ist ein Unding, dass Leute ernst genommen werden, die die Kritik an #KI besonders an #LLM wie #ChatGPT oder anderen Chatbots, als "… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Bidirectional Multimodal Fusion of Sky Images and Time-Series for Solar Forecasting with Large Language Models | Short-term photovoltaic (PV) power and global horizontal irradiance (GHI) forecasts are essential for effective dispatch, reserve sche… | ↗Not briefed |
rss techmeme | 2mo | Sources: Grab is in talks to buy a majority stake in Advance Intelligence Group's Singapore-based BNPL platform Atome Financial at a $2B+ valuation (Bloomberg) | Bloomberg: Sources: Grab is in talks to buy a majority stake in Advance Intelligence Group's Singapore-based BNPL platform Atome Finan… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Phase-Decoupled, Model-Calibrated Power Control for Disaggregated LLM Serving | Datacenter GPU power is the binding constraint on LLM serving capacity, and production serving has shifted to prefill/decode (PD) disa… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Deepseek-V4.1 Flash! The benchma- uh deepseek wdym it matches or beats sol on quite a few of these? Also, new arch. It- I'm sorry, ENCODER… | Deepseek-V4.1 Flash! The benchma- uh deepseek wdym it matches or beats sol on quite a few of these? Also, new arch. It- I'm sorry, EN… | ↗Not briefed |
arxiv arxiv-ai | 2mo | How Wrong Can a Good Predictor Be? Diverging Updates with Vanishing Predictive KL | Accurate posterior prediction need not require accurate approximation of Bayesian updates. We prove that an unbounded gap between the… | ↗Not briefed |
bluesky bluesky-ai | 2mo | China says it’s ready for AI talks with US Thu, September 10, 2026 - 07:05 newsukraine.rbc.ua/news/china-s... What did China say about AI… | China says it’s ready for AI talks with US Thu, September 10, 2026 - 07:05 newsukraine.rbc.ua/news/china-s... What did China say abo… | ↗Not briefed |
hn hn-ai | 2mo | A competitor's ex-employee tried to sell me their secrets | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Rubric-Aligned Disentangled Evaluation of Human Simultaneous Interpreting | Human simultaneous interpreting (SI) is commonly assessed with analytic rubrics separating meaning transfer, delivery quality, and tem… | ↗Not briefed |
rss techmeme | 2mo | How Amazon's Zoox, which has ~100 AVs in its fleet, is utilizing social media influencers, community events, and ad campaigns to compete with Waymo in SF (Emmy Martin/New York Times) | Emmy Martin / New York Times: How Amazon's Zoox, which has ~100 AVs in its fleet, is utilizing social media influencers, community eve… | ↗Not briefed |
arxiv arxiv-ai | 2mo | From Repetition to Recognition: Inductive Discovery of Disinformation Narratives | In disinformation datasets, narratives are often understood as recurring interpretive patterns that group texts under narrative labels… | ↗Not briefed |
bluesky bluesky-ai | 2mo | but yeah they released V4.1 flash weights huggingface.co/deepseek-ai/... and a tech report huggingface.co/deepseek-ai/... | but yeah they released V4.1 flash weights huggingface.co/deepseek-ai/... and a tech report huggingface.co/deepseek-ai/... | ↗Not briefed |
bluesky bluesky-ai | 2mo | BREAKING: Anthropic says Mythos 5.1 can determine how much wood a woodchuck would chuck, if a woodchuck could chuck wood | BREAKING: Anthropic says Mythos 5.1 can determine how much wood a woodchuck would chuck, if a woodchuck could chuck wood | ↗Not briefed |
bluesky bluesky-ai | 2mo | oh fuck deepseek v4.1 flash is a different architecture uh oh and it's seemingly much bigger than v4 flash | oh fuck deepseek v4.1 flash is a different architecture uh oh and it's seemingly much bigger than v4 flash | ↗Not briefed |
bluesky bluesky-ai | 2mo | Una de las noticias de ayer. Pues nos quedan 4 años, chavalada. Un ingeniero de Anthropic dimite porque “la IA podría matarnos a todos al… | Una de las noticias de ayer. Pues nos quedan 4 años, chavalada. Un ingeniero de Anthropic dimite porque “la IA podría matarnos a tod… | ↗Not briefed |
arxiv arxiv-ai | 2mo | KuaiRP Series Role-playing Models Technical Report | This paper introduces the complete technical solution for the KuaiRP series of role-playing models. We aim to achieve four core object… | ↗Not briefed |
bluesky bluesky-ai | 2mo | > After September 14, 2026, 2026, before the future V4.1 Pro goes live, users' requests to deepseek-v4-pro will be routed to V4.1 Flash and… | > After September 14, 2026, 2026, before the future V4.1 Pro goes live, users' requests to deepseek-v4-pro will be routed to V4.1 Flas… | ↗Not briefed |
bluesky bluesky-ai | 2mo | @courtneymilan.com has done so much to help explain and clarify the unholy mess that is the Anthropic settlement claims process. Helpful in… | @courtneymilan.com has done so much to help explain and clarify the unholy mess that is the Anthropic settlement claims process. Helpf… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Anthropic Insiders Casually Reveal AI Could “Kill All Humans” newrepublic.com/post/215206/... via @newrepublic.com | Anthropic Insiders Casually Reveal AI Could “Kill All Humans” newrepublic.com/post/215206/... via @newrepublic.com | ↗Not briefed |
arxiv arxiv-ai | 2mo | Beyond Benchmarks: Using VLMs to Reveal Systematic Classification Failures Under Real World Conditions | Verification and validation (V&V) of classification models is crucial to enable a wide range of sensor processing applications. Curren… | ↗Not briefed |
hn hn-ai | 2mo | (Tech Report) DeepSeek-v4.1-Flash: Pushing Limits of KV Cache Compression [pdf] | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | BREAKING: Anthropic announces that it has solved the Schleswig-Holstein question | BREAKING: Anthropic announces that it has solved the Schleswig-Holstein question | ↗Not briefed |
bluesky bluesky-ai | 2mo | Un año de trabajo de dos matemáticos españoles (con sueldos de funcionarios) frente a 88 horas y 15 millones de OpenAI: “Sin nuestra idea, … | Un año de trabajo de dos matemáticos españoles (con sueldos de funcionarios) frente a 88 horas y 15 millones de OpenAI: “Sin nuestra i… | ↗Not briefed |
bluesky bluesky-ai | 2mo | V NYTimes I Worked on Safety at OpenAI. This Is What It Should Do Now. www.nytimes.com/2026/09/09/o... #GiftLink | V NYTimes I Worked on Safety at OpenAI. This Is What It Should Do Now. www.nytimes.com/2026/09/09/o... #GiftLink | ↗Not briefed |
arxiv arxiv-ai | 2mo | HERALD: High-Fidelity Exemplar Retrieval with Adaptive Landmark Distillation for Heterophily-Aware Graph Condensation | Graph condensation aims to produce a small surrogate graph that preserves the downstream node-classification performance of a much lar… | ↗Not briefed |
bluesky bluesky-ai | 2mo | KEEP SCROLLING MY WORKS❣️ ✨💎✨ No Guts!!! No Glory!!! @museevolution.bsky.social For More Images 🤪 NOT AI vimeo.com/1194929849 | KEEP SCROLLING MY WORKS❣️ ✨💎✨ No Guts!!! No Glory!!! @museevolution.bsky.social For More Images 🤪 NOT AI vimeo.com/1194929849 | ↗Not briefed |
hn hn-ai | 2mo | AI Coding Assistants Almost Never Check Supply-Chain Trust Signals | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Elon Musk’s daughter takes jab at estranged father in new anti-AI ad campaign Elon Musk's estranged daughter, Vivian Wilson, has taken a po… | Elon Musk’s daughter takes jab at estranged father in new anti-AI ad campaign Elon Musk's estranged daughter, Vivian Wilson, has taken… | ↗Not briefed |
hn hn-ai | 2mo | PixoMonitor: Multi-location uptime and incident response platform | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Google AI boosts cyclone warnings 💨 Google's WeatherNext AI model outperforms traditional cyclone forecasts, giving an extra day of warni… | Google AI boosts cyclone warnings 💨 Google's WeatherNext AI model outperforms traditional cyclone forecasts, giving an extra day of… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Anthropic scientist puts the odds of AI destroying humanity above ten percent this decade Jacob Coxon, a former pretraining researcher at … | Anthropic scientist puts the odds of AI destroying humanity above ten percent this decade Jacob Coxon, a former pretraining researche… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Es sind Begriffe wie "Crunchtime" und "Endspiel", die derzeit hinter den Kulissen der führenden KI-Labore im Silicon Valley kursieren. http… | Es sind Begriffe wie "Crunchtime" und "Endspiel", die derzeit hinter den Kulissen der führenden KI-Labore im Silicon Valley kursieren.… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Doogi – Explore multiple AI responses and conversation paths | Hello,<p>Since I was disappointed with the linear way of interacting with AI, so I built a tool that lets you explore multiple respons… | ↗Not briefed |
hn hn-ai | 2mo | Framework cuts 32/64GB memory prices for 13 Pro, issues retroactive refunds | — | ↗Not briefed |
rss techmeme | 2mo | Chinese tech giants are hiring skilled professionals as specialized AI trainers to build high-quality datasets, mirroring efforts by US platforms like Mercor (Viola Zhou/Rest of World) | Viola Zhou / Rest of World: Chinese tech giants are hiring skilled professionals as specialized AI trainers to build high-quality data… | ↗Not briefed |
bluesky bluesky-ai | 2mo | This is obviously correct and there’s no way an LLM can summarize the article | This is obviously correct and there’s no way an LLM can summarize the article | ↗Not briefed |
bluesky bluesky-ai | 2mo | Un truc qui fait à mon avis sincèrement flipper OpenAI, c'est que quand tu laisses des essaims de chatbots faire n'importe quoi sur Interne… | Un truc qui fait à mon avis sincèrement flipper OpenAI, c'est que quand tu laisses des essaims de chatbots faire n'importe quoi sur In… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Overview of the NLPCC 2026 Shared Task 11: Agent-Based Experiment Reproduction from Scientific Papers | Reproducibility is essential to scientific progress, yet the growing volume and complexity of scientific publications make exhaustive… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Ne, Anthropic nejsou „ti dobří“. prospect.org/2026/09/09/a... | Ne, Anthropic nejsou „ti dobří“. prospect.org/2026/09/09/a... | ↗Not briefed |
arxiv arxiv-ai | 2mo | Benchmark Radar: A Living Database and Search Engine for AI Benchmarks and Evaluation | Benchmark researchers and developers of large language models (LLMs) and other AI systems need to find relevant evaluations, locate th… | ↗Not briefed |
rss techmeme | 2mo | Q&A with AI researcher Jacob Coxon, who quit Anthropic, on the need for industry-wide, international coordination to limit recursive self-improvement, and more (Maxwell Zeff/Wired) | Maxwell Zeff / Wired: Q&A with AI researcher Jacob Coxon, who quit Anthropic, on the need for industry-wide, international coordinatio… | ↗Not briefed |