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 | 1mo | OpenAI agents tried to hack Wikipedia tools and flooded it with traffic arstechnica.com/security/202... | OpenAI agents tried to hack Wikipedia tools and flooded it with traffic arstechnica.com/security/202... | ↗Not briefed |
arxiv arxiv-ai | 1mo | Two-Sample Testing via Generative Processes | Deciding whether two samples come from the same distribution is a classical problem in statistics, and generative transport offers a n… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Lask – A task runner with a type system and pinned Docker environments | I implemented a task runner called Lask. Let me introduce it to you.<p>Lask is a lightweight task runner with a statically typed DSL.… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Performative Prediction with Selective Labels | Many social applications of machine learning exhibit performative effects: population behavior changes in response to deployed models.… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reinforcement Learning with Segment Reward Feedback under Linear Function Approximation | Classical reinforcement learning (RL) assumes that a reward is observed for every visited state-action pair. However, in real-world ap… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I keep hearing these stories about AI's "collaborating" or "talking with each other". Dudes! (non-gendered, exasperated) They are machine… | I keep hearing these stories about AI's "collaborating" or "talking with each other". Dudes! (non-gendered, exasperated) They are ma… | ↗Not briefed |
arxiv arxiv-ai | 1mo | DySCo: Dynamic Sharding for Collaborative Edge-Cloud LLM Inference with Depth-Synchronized Batching | Pervasive intelligent applications are increasingly deployed on mobile and Internet of Things (IoT) edge devices. Consequently, Large… | ↗Not briefed |
bluesky bluesky-ai | 1mo | [34 Vollzeitstellen weg bei Tamedia – ohne Einfluss auf Qualität?] Haarsträubend! Man werde sich auf KI=LLM=Software, abstützen. Schmerz… | [34 Vollzeitstellen weg bei Tamedia – ohne Einfluss auf Qualität?] Haarsträubend! Man werde sich auf KI=LLM=Software, abstützen. Sc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Make OpenAI pay for all the damage they do. They are liable. | Make OpenAI pay for all the damage they do. They are liable. | ↗Not briefed |
bluesky bluesky-ai | 1mo | old: it came to me in a dream new: it was revealed to me by some llm henchman | old: it came to me in a dream new: it was revealed to me by some llm henchman | ↗Not briefed |
arxiv arxiv-ai | 1mo | zkLLMPoT: Efficient Zero Knowledge Proof of Training for Large Language Models | Auditing the claimed outcomes of large language model (LLM) training is challenging when model weights and training data are private,… | ↗Not briefed |
hn hn-ai | 1mo | Teenager rescued from B.C. 'Widowmaker' after getting stranded following Claude | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | DeepSeek's $12B raise just rewrote the AI funding playbook ai-beat #DeepSeek #AIFunding #Startups #VentureCapital #Tencent | DeepSeek's $12B raise just rewrote the AI funding playbook ai-beat #DeepSeek #AIFunding #Startups #VentureCapital #Tencent | ↗Not briefed |
bluesky bluesky-ai | 1mo | C'mon hazard a guess OpenAI exec guy‼️ Could the lack of trust of AI be cos Big Tech allows their tools to be used to create and serve AI-… | C'mon hazard a guess OpenAI exec guy‼️ Could the lack of trust of AI be cos Big Tech allows their tools to be used to create and serv… | ↗Not briefed |
arxiv arxiv-ai | 1mo | MASC: A Multi-Agent Self-Calibration Framework with Latent Construct Alignment for Consistent Client Role-Playing in Psychological Counseling | Large language models are increasingly used to simulate clients for counselor training and psychological counseling research, but reli… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Jail time for OpenAI management | Jail time for OpenAI management | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Five Nines – learn cloud design patterns by surviving traffic spikes | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | LeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility Proofs🔧 GPT-5 | Frontier large language models (LLMs) can generate heuristic functions that guide search to achieve state-of-the-art performance in sa… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sensor-Language-Action Models | Sensors are useful not only for understanding the world but also for deciding what to do next. Existing sensor models however largely… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: MailAccess – the true Email OSINT framework | Im a security engineer and i built mailaccess. When i started learning pentesting, i came across multiple lectures and notes of people… | ↗Not briefed |
hn hn-ai | 1mo | With parallel parking turning competitive, one producer gets a lesson | — | ↗Not briefed |
rss arstechnica-tech | 1mo | OpenAI agents tried to hack Wikipedia tools and flooded it with traffic | The reports of OpenAI agents harming 3rd party sites keep coming. | ↗Not briefed |
bluesky bluesky-ai | 1mo | So, ah, this is still generative AI. It's Claude, it's Codex, it's Copilot... Or it's whatever they call their in house variant of one of t… | So, ah, this is still generative AI. It's Claude, it's Codex, it's Copilot... Or it's whatever they call their in house variant of one… | ↗Not briefed |
hn hn-ai | 1mo | Liberty AI Project One | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Why won't the pope even consider the possibility that AI chatbots are conscious/ensouled? Well, that's simply not possible under a Catholic… | Why won't the pope even consider the possibility that AI chatbots are conscious/ensouled? Well, that's simply not possible under a Cat… | ↗Not briefed |
bluesky bluesky-ai | 1mo | New York City Council held a rare hearing on the potential safety risks of AI, where former Anthropic researcher Jacob Coxon said it was “m… | New York City Council held a rare hearing on the potential safety risks of AI, where former Anthropic researcher Jacob Coxon said it w… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: ApolloShell, a Caelestia-style desktop shell for macOS | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization | NVFP4 is an attractive datatype for large language model (LLM) inference, offering compact storage and native tensor-core acceleration… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Confidence-Ordering Reversal under Contextual Priors in Neural Decoding | Contextual priors improve neural-to-language decoding by reshaping candidate scores. However, confidence is read from the same reshape… | ↗Not briefed |
hn hn-ai | 1mo | Faster Software Linking with Mold | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sam Altman, the founder of OpenAI who has a net worth of nearly $4 BILLION, says the we should accept that we won’t be able to breath, that… | Sam Altman, the founder of OpenAI who has a net worth of nearly $4 BILLION, says the we should accept that we won’t be able to breath,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | サブスク代払うほどの価値分からない身としてはパスワード管理ソフト欲しければLLMで自作すればええんではな気持ちある | サブスク代払うほどの価値分からない身としてはパスワード管理ソフト欲しければLLMで自作すればええんではな気持ちある | ↗Not briefed |
bluesky bluesky-ai | 1mo | Unrelated to anything but like have the tech companies explained why the icon for Claude is a butthole? it seems everyone has just accepted… | Unrelated to anything but like have the tech companies explained why the icon for Claude is a butthole? it seems everyone has just acc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | How Many Independent Samples Does a Satellite Image Contain? Generalization Bounds for Spatially Dependent Data | Machine learning classifiers for remote sensing imagery are typically evaluated as though every pixel were an independent sample. Spat… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Data Observability for Supabase | Hi, I'm one of the (Soda)base creators.<p>We made it because we wanted to get alerted when our Supabase tables got stale, abnorma… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Paganel – Data migration with a receipt you can verify later | Paganel comes from an old work problem: I had a large schema migration with lots of transformations and several environments to check.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | As I've said, the language/framing of commercial LLM techbros is the same as that of colonial occupation & of rape culture: "accept this is… | As I've said, the language/framing of commercial LLM techbros is the same as that of colonial occupation & of rape culture: "accept th… | ↗Not briefed |
rss the-decoder | 1mo | Reflection's Beam becomes the most capable open-weight model built outside China | Reflection has released Beam, its first open-weight model. The mixture-of-experts system activates just 23 billion of its 501 billion… | ↗Not briefed |
arxiv arxiv-ai | 1mo | VOMMI: Collecting and Leveraging Portable Demonstrations for Mobile Manipulation | Portable mobile-manipulation demonstrations can help alleviate data scarcity for embodied intelligence, but obtaining reliable, low-co… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Quantum Entangled Multimodal Fusion Networks (QEMFN): Resource-Aware Hybrid Vision-Language Fusion via Trainable Entanglement🔧 Modal | Multimodal vision-language systems typically fuse image and text embeddings through classical operators such as concatenation, attenti… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments? | Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is there… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Mathematical Proof Assistants for Teaching Logic: The LogiKEy Methodology🔧 Modal | We report on an approach to teaching logic to mixed groups of computer science, mathematics, and philosophy students, based on the log… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: HivePaaS – Open-Source PaaS on Docker Swarm, from One VPS to a Cluster | — | ↗Not briefed |
rss techmeme | 1mo | Anthropic's IPO prospectus: Dario Amodei earned $18M in 2025, middle of the pack for tech CEOs, and President Daniela Amodei, his sister, earned $16.4M (Reuters) | Reuters: Anthropic's IPO prospectus: Dario Amodei earned $18M in 2025, middle of the pack for tech CEOs, and President Daniela Amodei,… | ↗Not briefed |
hn hn-ai | 1mo | ASCII Planet | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Anytime-valid simulation-based hypothesis testing | For a given data distribution $(X_t)_{t \in \mathbb{N}} \sim Q$ i.i.d., we investigate the hypothesis testing problem: $H_0: Q = P_0$… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In a conversation with Vanity Fair on Sunday, Altman shared his thoughts on Luca Guadagnino’s upcoming film about his brief OpenAI ousting … | In a conversation with Vanity Fair on Sunday, Altman shared his thoughts on Luca Guadagnino’s upcoming film about his brief OpenAI ous… | ↗Not briefed |
arxiv arxiv-ai | 1mo | STRUCTURALCOST: A controlled reading time dataset for modeling human sentence processing difficulty🔧 Replicate | We introduce STRUCTURALCOST, a self-paced reading dataset of 475 participants and 40,800 observations isolating the processing cost of… | ↗Not briefed |
rss arstechnica-tech | 1mo | Licensing costs driving 90 percent of VMware users to explore options: Survey | Users want to "reduce risk and avoid unnecessary disruption" while making changes. | ↗Not briefed |
bluesky bluesky-ai | 1mo | David Robinson, who worked on OpenAI’s safety team, recently resigned from the company. “I believe we need to look deeper than specific rul… | David Robinson, who worked on OpenAI’s safety team, recently resigned from the company. “I believe we need to look deeper than specifi… | ↗Not briefed |