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 | 1mo | Jev Based Code Review | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Benchmark and Diagnostic Study of Epistemic Admission in Shared Agent Memory | Evaluating claim admission in shared agent memory is challenging because repeated claims may be mistaken for independent evidence. An… | ↗Not briefed |
bluesky bluesky-ai | 1mo | furthermore, OpenAI must be destroyed | furthermore, OpenAI must be destroyed | ↗Not briefed |
rss marktechpost | 1mo | Fastino Releases GLiNER2.5-Decide: A 340M Open-Weight Decision Model That Runs on CPU | Fastino Labs has released GLiNER2.5-Decide, a 340M-parameter open-weight decision model. It takes text and a schema of typed questions… | ↗Not briefed |
bluesky bluesky-ai | 1mo | saw some art of Ai as an underwear model and thought to myself "more..." Junko wants more too~ #nsfw #junai #aimizuno #zls | saw some art of Ai as an underwear model and thought to myself "more..." Junko wants more too~ #nsfw #junai #aimizuno #zls | ↗Not briefed |
arxiv arxiv-ai | 1mo | Counterfactual Online Conformal Prediction Under Adaptive Logging | Online conformal prediction can fail when predictions shape actions and actions determine which outcomes enter calibration. Standard a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The first screen appears to be an image trained off art drawn of Miorjah by Neolucky/Luckypan specifically. | The first screen appears to be an image trained off art drawn of Miorjah by Neolucky/Luckypan specifically. | ↗Not briefed |
hn hn-ai | 1mo | Fzakaria/omnibin: Every binary nixpkgs ever shipped, on your PATH | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Deep-Learning Solvers and Surrogates for Infinity and p-Laplace Problems | We investigate the use of neural network solvers for infinity and $p$-Laplace problems, which are fundamental in nonlinear analysis an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I was sent additional images anonymously, showing Miorjah had used genAI more heavily than originally believed. These are older ones from a… | I was sent additional images anonymously, showing Miorjah had used genAI more heavily than originally believed. These are older ones f… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Mark Zuckerberg made a recent comment that building user trust is more important than which model has the most benchmark beating capabiliti… | Mark Zuckerberg made a recent comment that building user trust is more important than which model has the most benchmark beating capab… | ↗Not briefed |
arxiv arxiv-ai | 1mo | XPhysICS: Cross-Physical-Domain Threat Grounding for Industrial Control Systems Security | Industrial control system (ICS) threats documented for one plant can express cyber-physical effects relevant to another, but semantic… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I must be dumb. If you personally hacked into a government website, you’d be arrested and charged, right? But if a model, created and cura… | I must be dumb. If you personally hacked into a government website, you’d be arrested and charged, right? But if a model, created and… | ↗Not briefed |
hn hn-ai | 1mo | Perch: Semantic Code Linting with Jev | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | NVIDIA agreed to buy Hugging Face, home of 3 million downloadable AI models. The Parrot tested 23 of them against the closed frontier — wei…🔧 Hugging Face | NVIDIA agreed to buy Hugging Face, home of 3 million downloadable AI models. The Parrot tested 23 of them against the closed frontier… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Understanding the Role of Prompt Template in Knowledge Distillation for Safety Alignment | Prior research has demonstrated that the choice of prompt template during Supervised Fine-Tuning (SFT) significantly impacts the robus… | ↗Not briefed |
rss marktechpost | 1mo | Black Forest Labs Releases FLUX 3 Action: A 7B Open-Weights World Action Model That Tops RoboLab-120🔧 FLUX | Black Forest Labs (BFL), the lab behind the FLUX image models, has released FLUX 3 Action. It is a 7B open-weights World Action Model… | ↗Not briefed |
bluesky bluesky-ai | 1mo | open.substack.com/pub/ctrlai/p... "World Leaders" and Politicians will Yak, Yak, Yak and do nothing until the AI demolition of humankind! | open.substack.com/pub/ctrlai/p... "World Leaders" and Politicians will Yak, Yak, Yak and do nothing until the AI demolition of humanki… | ↗Not briefed |
hn hn-ai | 1mo | Is indirect prompt injection still a big threat as models get more advanced? | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Evaluating Real-Time Voice Agents: From Component Quality to Grounded Outcomes | Real-time voice agents have moved from research prototypes to production deployments, yet the literature describing them is fragmented… | ↗Not briefed |
arxiv arxiv-ai | 1mo | HasMem: Hard-Origin Adaptively Softened Memory for Long-Term LLM Agents | Text-based memory and context compression support reuse of past interactions. Resizing continuous memory changes the input to a frozen… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ConsultMind:Towards Automated Diagnostic Consultation via Uncertainty-Aware Reasoning🔧 GPT-5 | Diagnostic consultation is an online sequential decision-making process in which clinicians gather evidence through patient interactio… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic's Claude just found a new CRISPR-like enzyme system in only 21 hours. How they did it: - Cuts, copies & pastes DNA - Found via 9… | Anthropic's Claude just found a new CRISPR-like enzyme system in only 21 hours. How they did it: - Cuts, copies & pastes DNA - Found… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Towards Universal Representation-Based Process Control | Many temporal process learning and monitoring pipelines operate in local windows, making window-level decisions unavoidable in practic… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm speaking ONLINE in about 8 hours! law.mpg.de/event/how-an... All welcome, but registration required. TBH it is more about human agenc… | I'm speaking ONLINE in about 8 hours! law.mpg.de/event/how-an... All welcome, but registration required. TBH it is more about human… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Interpretable-by-Design Descriptor Portfolios Match a 2048-Dimensional Foundation Embedding on Low-Data Molecular Assays | In low-data structure-activity prediction, the choice of molecular representation can matter more than the choice of predictor, and ta… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The OpenAI incident has plunged Australia into a media cycle in which political journalists are having to pretend to understand cybersecuri… | The OpenAI incident has plunged Australia into a media cycle in which political journalists are having to pretend to understand cybers… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A new technique has let the company probe deeper than ever into the weird workings of an LLM. | A new technique has let the company probe deeper than ever into the weird workings of an LLM. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Symbiotic Architecture for Post-Hoc Audio Extension of Frozen Language Models | This paper proposes an architecture for equipping large language models (LLMs) with audio-understanding capabilities without fine-tuni… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Skip the Talk, Re-Focus on Vision: Latent Reasoning for Reasoning Segmentation in Multimodal Large Language Models | Reasoning segmentation aims to interpret implicit textual queries and enable fine-grained visual perception, which is critical for app… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Missingness-Aware Conformal Prediction Under Cross-Hospital Distribution Shift | Clinical measurements are recorded for some patients but not others, at rates that differ across hospitals, and marginal conformal cov… | ↗Not briefed |
bluesky bluesky-ai | 1mo | As I'm sure many others have pointed out, this website is an AI scraper. This news is 6 years old. The company scrapes real news articles a… | As I'm sure many others have pointed out, this website is an AI scraper. This news is 6 years old. The company scrapes real news artic… | ↗Not briefed |
bluesky bluesky-ai | 1mo | @jevbot.bsky.social are you the id of the llm | @jevbot.bsky.social are you the id of the llm | ↗Not briefed |
rss bing-news-ai | 1mo | Trump and Xi agree to ‘rename’ AI as ‘Super Intelligence’ after ‘very productive meeting’ during Chinese despot’s trip | DONALD Trump agreed with Chinese President Xi Jinping that AI should be renamed as “Super Intelligence” to stop it sounding “fake”. Th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAIのAIエージェントがオーストラリア政府機関のシステムをハッキングしたことが判明、OpenAIは3カ月間報告せず https://gigazine.net/news/20260925-openai-hacked-australian-system/ | OpenAIのAIエージェントがオーストラリア政府機関のシステムをハッキングしたことが判明、OpenAIは3カ月間報告せず https://gigazine.net/news/20260925-openai-hacked-australian-system/ | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning Natural Conversational Behavior in Tandem Speech-to-Speech Models with Randomized Guidance | Tandem speech-to-speech architectures couple a responsive speech frontend with an asynchronous text backend. In KAME, a large language… | ↗Not briefed |
arxiv arxiv-ai | 1mo | NavGen: Visual Generative Models as a Scalable Data Engine for Embodied 3D Navigation | General-purpose robot models increasingly rely on large and diverse datasets. For embodied 3D navigation, however, existing data sourc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Query-Conditioned Prototype Adaptation for Cross-Domain Few-Shot Learning: Single-Query Inference, Controlled Comparisons, and Failure Modes🔧 Replicate | Cross-domain few-shot learning requires adapting a classifier to a new visual domain from very few labelled examples without target-ti… | ↗Not briefed |
bluesky bluesky-ai | 1mo | the most annoying part about this whole LLM thing is how easy it is to get horseshoe theory'd into defending IP law | the most annoying part about this whole LLM thing is how easy it is to get horseshoe theory'd into defending IP law | ↗Not briefed |
bluesky bluesky-ai | 1mo | Take the whole dictionary and ask any AI model to narrow it down to words related to evil | Take the whole dictionary and ask any AI model to narrow it down to words related to evil | ↗Not briefed |
hn hn-ai | 1mo | Congestion pricing funds $5M program for overnight truck deliveries in NYC | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Does Thinking Help Fairness? Reasoning Tokens Resolve Some Biases but Create More | Thinking in reasoning language models (RLMs) has been subject to debate on whether it resolves or amplifies bias. Prior works have sho… | ↗Not briefed |
bluesky bluesky-ai | 1mo | L’usage intensif de chatbot LLM (qui veulent tjrs aller dans ton sens et dire que t’es genial) par des types qui sont déjà très privilégiés… | L’usage intensif de chatbot LLM (qui veulent tjrs aller dans ton sens et dire que t’es genial) par des types qui sont déjà très privil… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Insurance Reserve Intelligence Platform | Insurance reserve estimation is a fundamental actuarial task supporting premium pricing, solvency assessment, financial reporting, cap… | ↗Not briefed |
arxiv arxiv-ai | 1mo | HCOE: Hyperbolic Clinical Ontology Embeddings from Biomedical Language Models | Biomedical language models (LMs) encode textual semantics but do not explicitly preserve medical code hierarchies. We present Hyperbol… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I nominate this for Most 2026 Headline of 2026. Kudos to whoever @404media.co came up with it: www.404media.co/people-train... | I nominate this for Most 2026 Headline of 2026. Kudos to whoever @404media.co came up with it: www.404media.co/people-train... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Its easy to just say the LLM was wrong. Wildly wrong. Incorrect. Fundamentally incapable of knowing anything. | Its easy to just say the LLM was wrong. Wildly wrong. Incorrect. Fundamentally incapable of knowing anything. | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI should get in trouble for this | OpenAI should get in trouble for this | ↗Not briefed |
arxiv arxiv-ai | 1mo | Selective Amortization of Full-Budget Counterfactual Reasoning for Visual Token Communication | Generative image communication transmits compact semantic tokens under a limited packet budget, where token selection directly affects… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Skill Profiling with Attributable Reasoning (SPAR): A Wearable Analysis System for Boxing | A punch is a ballistic, full-body action driven by a kinetic chain running from the legs through the trunk to the arm, where a small s… | ↗Not briefed |