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 | 2mo | Ursula von der Leyen has name-checked the incident where OpenAI agents hacked HuggingFace this summer while discussing AI development. Mor… | Ursula von der Leyen has name-checked the incident where OpenAI agents hacked HuggingFace this summer while discussing AI development.… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Yes but I can tell you're not an LLM because you didn't respond with "great question!" | Yes but I can tell you're not an LLM because you didn't respond with "great question!" | ↗Not briefed |
arxiv arxiv-ai | 2mo | Building Trust in Artificial Intelligence: A Necessity for Railway Applications | Artificial Intelligence (AI) is currently only applied to non-safety critical applications due to the strict standards and regulations… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Don't believe the AI Tech Bros about slowing down it's a trick to manipulate the markets and for their spending power for chips. China is … | Don't believe the AI Tech Bros about slowing down it's a trick to manipulate the markets and for their spending power for chips. Chin… | ↗Not briefed |
arxiv arxiv-ai | 2mo | I code or AI code: A comparative evaluation of AI-rated scores in classroom observations🔧 GPT-5 | Classroom observations are widely recognized as a key tool for establishing benchmarks of education quality and guiding pedagogical im… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Behavioral Fingerprinting and Navigation Prediction in Web Browsing | Web browsing often appears ephemeral: users visit a few websites, complete a task, and move on. However, even short fragments of brows… | ↗Not briefed |
hn hn-ai | 2mo | Gemini for Go Developers: Building Agents in Go | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Who Audits Whom, on What Substrate, with What Evidence? An Independence-Graded Audit Protocol for Agentic AI | Agentic AI systems plan, invoke tools and act with limited supervision; they are now both the subject of audits and, increasingly, the… | ↗Not briefed |
hn hn-ai | 2mo | Omarchy ships its own Linux kernel in v4.0.4 | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | BENCHCOMPASS: From Scores to Signals for Training and Harness Decisions in Payment-Domain LLMs | Payment operations are a critical financial infrastructure, but the value of large language models in this domain remains unclear beca… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Are people that dumb? Are they really falling for this? "We're the only ones who can save you from our terrible creation, so please keep se… | Are people that dumb? Are they really falling for this? "We're the only ones who can save you from our terrible creation, so please ke… | ↗Not briefed |
hn hn-ai | 2mo | Never-again – so your AI agent stops repeating mistakes you fixed | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | I'm betting they also read LLM-"summaries" of, or LLM-"arguments" against, my work. What these "summaries" & "arguments" nicely illustrate… | I'm betting they also read LLM-"summaries" of, or LLM-"arguments" against, my work. What these "summaries" & "arguments" nicely illus… | ↗Not briefed |
bluesky bluesky-ai | 2mo | LLMs turn coal into theorems. Or, more precisely: OpenAI turns one panamax bulk carrier of coal into Navier-Stokes. | LLMs turn coal into theorems. Or, more precisely: OpenAI turns one panamax bulk carrier of coal into Navier-Stokes. | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI chatbots developed a secret language that baffled humans, study says | AI chatbots developed a secret language that baffled humans, study says | ↗Not briefed |
arxiv arxiv-ai | 2mo | REPAIR: Resolving Long-Tail Confusion in Scientific Retrievers via Fact-Verified Iterative Refinement | Precise retrieval of scientific information is fundamentally constrained by long-tailed concepts and high fact-sensitivity of scientif… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The inability to genuinely engage on substance is a consistent sign of heavy LLM-users. It is interesting and perhaps telling that the OP s… | The inability to genuinely engage on substance is a consistent sign of heavy LLM-users. It is interesting and perhaps telling that the… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Obviously, I entirely trust OpenAI and Nvidia bosses to police and manage the threat of AI. How staggering that people could think these bi… | Obviously, I entirely trust OpenAI and Nvidia bosses to police and manage the threat of AI. How staggering that people could think the… | ↗Not briefed |
arxiv arxiv-ai | 2mo | ${M}^2$Tok: Multi-head Multi-codebook Discrete Action Tokenization for Vision-Language-Action Models | Recent advancements have successfully adapted autoregressive language models to process multimodal signals, such as images and actions… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 난 솔직하게 LLM보다 초파리 뇌로 한국어 로컬 모델 돌리는게 더 무섭던데. | 난 솔직하게 LLM보다 초파리 뇌로 한국어 로컬 모델 돌리는게 더 무섭던데. | ↗Not briefed |
bluesky bluesky-ai | 2mo | It should be obvious, but the only way an LLM destroys humanity is if the humans plug it into human systems that could likewise destroy hum… | It should be obvious, but the only way an LLM destroys humanity is if the humans plug it into human systems that could likewise destro… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Warum kauft man überhaupt einen Roboter wo KI verwendet (also wenn's um ein LLM geht und es nicht einfach ein fancy Wort für komplexen Algo… | Warum kauft man überhaupt einen Roboter wo KI verwendet (also wenn's um ein LLM geht und es nicht einfach ein fancy Wort für komplexen… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Re2A: Situated Conversational Recommendation via Rubric-based Preference Reasoning and Alignment | Real-world recommendation scenarios are commonly grounded in shared physical environments during user-recommender interactions. This m… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Quanta: A Self-Contained Python Library for Hybrid Retrieval over Quantised Embeddings, Lexical Indexes, and Knowledge Graphs | An advanced retrieval-augmented generation pipeline is typically assembled from three or four independently operated systems: an appro… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Hmm, let's think about their track record and how trustworthy that makes them ... | Hmm, let's think about their track record and how trustworthy that makes them ... | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Point at anything in your app, say what you want, get a spec | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: LaunchPad – I built a Launchpad to replace macOS Apps | Download URL: <a href="https://api.ihave.website/launchpad/download/macos" rel="nofollow">https://a… | ↗Not briefed |
hn hn-ai | 2mo | UK mathematician debunks myth around Parthenon's optical illusions | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Competition for AI dominance resembles the nuclear arms race of the Cold War era. A similarly serious response may also now be needed. | Competition for AI dominance resembles the nuclear arms race of the Cold War era. A similarly serious response may also now be needed. | ↗Not briefed |
arxiv arxiv-ai | 2mo | Acting in Meters: Learning Metric Interactions for Precise Robotic Manipulation | Vision-Language-Action models and World-Action Models have advanced language-conditioned robotic manipulation, yet often leave metric… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Not even errors, more a statistical certainty. 'Error' implies a fault that can be fixed. All the LLM builders can do is try to hone them s… | Not even errors, more a statistical certainty. 'Error' implies a fault that can be fixed. All the LLM builders can do is try to hone t… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Trust tech-bosses to do the right thing? Let Sam Altman regulate himself? Er, no, hard pass, we have actually seen what you‘re like Sam. ww… | Trust tech-bosses to do the right thing? Let Sam Altman regulate himself? Er, no, hard pass, we have actually seen what you‘re like Sa… | ↗Not briefed |
arxiv arxiv-ai | 2mo | F-DACE: Fuzzy Disagreement-Aware Causal Evidence Fusion for Abstention-Safe Conversational Retail Decision Support🔧 LangGraph | Observational decision-support systems often expose one causal estimate as a recommendation even when plausible estimators disagree. T… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Alors qu'on discute des risques de l'IA (sans en regarder la hiérarchie ni la distribution), les acteurs de la tech organisent le champ de … | Alors qu'on discute des risques de l'IA (sans en regarder la hiérarchie ni la distribution), les acteurs de la tech organisent le cham… | ↗Not briefed |
hn hn-ai | 2mo | Ask HN: How do you test an app from a region you're not physically in?🔧 Replicate | We run a small B2B product for a few overseas markets. We keep seeing bugs that only reproduce locally: country-specific payment SDK b… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Now we know they're using LLMs to assess we'll be forced to run proposals through LLMs before submission and making changes to improve scor… | Now we know they're using LLMs to assess we'll be forced to run proposals through LLMs before submission and making changes to improve… | ↗Not briefed |
bluesky bluesky-ai | 2mo | E pa to me zanima. Koji misaoni proces dovede do toga da se koristi LLM za izradu web stranice, ali "kako napraviti instagram" je pitanje z… | E pa to me zanima. Koji misaoni proces dovede do toga da se koristi LLM za izradu web stranice, ali "kako napraviti instagram" je pita… | ↗Not briefed |
bluesky bluesky-ai | 2mo | My understanding is that the evidence shows LLM-based assessment gives better 'marks' to LLM-written things (essays, grants etc). So by cho… | My understanding is that the evidence shows LLM-based assessment gives better 'marks' to LLM-written things (essays, grants etc). So b… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Anomaly Detection in General Ledger Data: Results from a Hybrid Approach | Journal Entry Tests (JETs) are a mandatory part of annual audits to evaluate and assess both highrisk audit areas and potential materi… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 【「監視されていないときのAIは、もう評価できない」】 米OpenAIの研究者、ダニエル・セルサム氏は9月14日(現地時間)、AIのリスクに関する個人的な見解をまとめた声明をGoogleドキュメントで公開した。同氏はXアカウントを持たないため、元OpenAI研究者でAI Fut… | 【「監視されていないときのAIは、もう評価できない」】 米OpenAIの研究者、ダニエル・セルサム氏は9月14日(現地時間)、AIのリスクに関する個人的な見解をまとめた声明をGoogleドキュメントで公開した。同氏はXアカウントを持たないため、元OpenAI研究者でA… | ↗Not briefed |
hn hn-ai | 2mo | Ask HN: How do you get coding agents to clean up after changing direction? | I've noticed the same failure mode across most coding agents.<p>When an implementation runs into a problem, such as a function no… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Got a good 96gb-VRAM local LLM harness going and can't believe I used to just... put up with the garbage latencies that OpenAI/Anthropic/Go… | Got a good 96gb-VRAM local LLM harness going and can't believe I used to just... put up with the garbage latencies that OpenAI/Anthrop… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Remembering Solomon Marcus | From the manifest of Andre Breton, through the transdisciplinary understanding, we arrive at a post-modern manifest. A talk by Laura D… | ↗Not briefed |
bluesky bluesky-ai | 2mo | This is the worst possible outcome. We are going to drive this fucking thing with gas. Ruining the environment like is happening in USA now… | This is the worst possible outcome. We are going to drive this fucking thing with gas. Ruining the environment like is happening in US… | ↗Not briefed |
rss marktechpost | 2mo | Prior Labs Releases TabPFN-3.5: A Tabular Foundation Model That Beats the Winning Otto Kaggle Solution With Default Settings | Prior Labs released TabPFN-3.5, a tabular foundation model pretrained only on synthetic data that beats Otto's winning solution. The p… | ↗Not briefed |
arxiv arxiv-ai | 2mo | APGEM: Adaptive Policy-Guided Error Mitigation for Quantum Reinforcement Learning on a Real-World CVRP Case Study | Quantum Reinforcement Learning (QRL) represents policies as variational quantum circuits (VQCs), making it attractive for combinatoria… | ↗Not briefed |
hn hn-ai | 2mo | Bootstrap from Twitter | — | ↗Not briefed |
hn hn-ai | 2mo | GPT Image 2.5 Prompts | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | CPR: Combining global composing, local performing and full-sequence refining in piano rendering with continuous autoregressive modelling | Prompt-conditioned piano MIDI-to-Music rendering aims to faithfully render target notes while reproducing the timbre of a reference re… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The Discovery When an artificial intelligence model answers a clinical question about a chest X-ray, it seems obvious that it is looking at… | The Discovery When an artificial intelligence model answers a clinical question about a chest X-ray, it seems obvious that it is looki… | ↗Not briefed |