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 | Ask HN: Will AI Longevity Models Optimize for Physical Scarcity? | Since longevity has a high value and AI models are becoming capable of new leaps I think the dynamic of the market might look for phys… | ↗Not briefed |
hn hn-ai | 1mo | We ran 5 LLMs 100 times each against tool calls that time out after the write | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | A teenager in Canada took a gondola up a mountain and asked Claude to plot him a hike down. He ended up stranded on a sheer ledge and had t… | A teenager in Canada took a gondola up a mountain and asked Claude to plot him a hike down. He ended up stranded on a sheer ledge and… | ↗Not briefed |
arxiv arxiv-ai | 1mo | What the Sleeve Feels: Explainable Machine Learning for Textile Pressure-Based Postural Screening | Pressure-sensing smart textiles convert body-surface contact into a dense, image-like signal closely tied to posture and movement, mak… | ↗Not briefed |
bluesky bluesky-ai | 1mo | surely tao is savvy enough to realize that openai places infinitesimal value on the health of mathematics compared to the prospects of futu… | surely tao is savvy enough to realize that openai places infinitesimal value on the health of mathematics compared to the prospects of… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Drosophila Whole-Connectome Network Can Learn Human-Designed Cognitive Tasks | Can a biological wiring diagram serve as a useful computational substrate beyond the behaviors for which it evolved? We use the public… | ↗Not briefed |
hn hn-ai | 1mo | Argos –- A 1MB Rust Security Shim for Model Context Protocol Servers | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | These will totally not be written by LLM | These will totally not be written by LLM | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: How do you use and interpret OpenAI Decisions API | I've had some strange results with Jev, and I question the "probabilistic" aspects of it. So I put OpenAI Decisions API… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Huh Went to grad school and law school with David. Will be interested in reading this. www.nytimes.com/2026/10/07/o... | Huh Went to grad school and law school with David. Will be interested in reading this. www.nytimes.com/2026/10/07/o... | ↗Not briefed |
rss techmeme | 1mo | Stuut, which uses AI agents to automate enterprise order-to-cash cycles and serves 150+ clients, raised a $52.5M Series B led by Insight, for $93M total funding (Mike Wheatley/SiliconANGLE) | Mike Wheatley / SiliconANGLE: Stuut, which uses AI agents to automate enterprise order-to-cash cycles and serves 150+ clients, raised… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Having recently attended a 4-hour "AI Essentials for Parents" session at AFT's National Academy for AI Instruction, I can confirm! It began…🔧 ChatGPT | Having recently attended a 4-hour "AI Essentials for Parents" session at AFT's National Academy for AI Instruction, I can confirm! It… | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLMはサボるし劣化するし、人間の監視なしに作業を任せてはならない｜地雷魚 note.com/jiraygyo/n/n... ここで解説しているけど、LLMは技術的特性から「根本的に反復したタスクに向いていない」というものがあり、1%のミスが指数関数的に累積していくので、絶対… | LLMはサボるし劣化するし、人間の監視なしに作業を任せてはならない｜地雷魚 note.com/jiraygyo/n/n... ここで解説しているけど、LLMは技術的特性から「根本的に反復したタスクに向いていない」というものがあり、1%のミスが指数関数的に累積していく… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Beyond Reward Suppression: Near-Optimal Offline Attacks on Warm-Start Bandits with Bounded Rewards | Adversarial attacks on bandits aim to mislead a learner toward a target arm while keeping the attack cost small. Existing attacks typi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Temporal Predictive Multiplicity: Equally Accurate Time Series Models Yield Different Forecast Trajectories | Models with near-identical predictive performance can yield substantially different predictions, a phenomenon known as predictive mult… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Efficient Patch-Based Anomaly Detection Fused with Diffusion Driven Generative Modeling for Semiconductor Wafer Bin Map Open Set Anomaly Detection | Spatial defect signatures on wafer bin maps (WBMs) trace yield loss to specific process faults, yet supervised classifiers recognize o… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "AI and digital surveillance reversed the American consumer model. We grew up with people constantly trying to sell us stuff. Now they're t… | "AI and digital surveillance reversed the American consumer model. We grew up with people constantly trying to sell us stuff. Now they… | ↗Not briefed |
bluesky bluesky-ai | 1mo | once you get used to generating Claudelish documentation for yourself, you see LLM-generated writing everywhere | once you get used to generating Claudelish documentation for yourself, you see LLM-generated writing everywhere | ↗Not briefed |
arxiv arxiv-ai | 1mo | Marrying Pricing and Advertising with LLMs | We study a sequential pricing problem in which a seller jointly posts a price and an advertisement generated by a large language model… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I haven't really felt the impact of the openAI math results they published. I'll probably start feeling it when people stop paying me for l… | I haven't really felt the impact of the openAI math results they published. I'll probably start feeling it when people stop paying me… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Extreme Binary Classification: Extreme Value Theory for Extreme Constraint on False Negative | While binary classification is one of the most extensively studied problems in machine learning, the regime in which the goal is to le… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A 50k line PR that the person didn't write (an LLM did) that they expect the maintainers to read, review, understand, and then put into the… | A 50k line PR that the person didn't write (an LLM did) that they expect the maintainers to read, review, understand, and then put int… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The UN AI panel's first Thematic Brief "sets a dangerous precedent at the science–policy interface," argue Jason Tucker, Virginia Dignum an…🔧 Hugging Face | The UN AI panel's first Thematic Brief "sets a dangerous precedent at the science–policy interface," argue Jason Tucker, Virginia Dign… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sensitive-Topic Leakage Through LLM Routing Metadata: Measurement and Mitigation | LLM routers pick a cheap or expensive model per request by its content, and many gateways and some cloud platforms can log that choice… | ↗Not briefed |
hn hn-ai | 1mo | Modern SVG flight instruments for React | — | ↗Not briefed |
rss huggingface-blog | 1mo | One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Procinsh – A 3D Linux process inspector | — | ↗Not briefed |
hn hn-ai | 1mo | Claude for Startups Program | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | 何度も書いてますが 今のLLM（大規模言語モデル、AI）って高性能VLM（視覚言語モデル）が必須に成ってます。 何故か？ 「画像を見て何かをする」 これ、高性能のVLMが無いと出来ない。 LLMが画像を認知する為にはVLMが必須。 だから、画像生成をやらない所まで本気でVLM開… | 何度も書いてますが 今のLLM（大規模言語モデル、AI）って高性能VLM（視覚言語モデル）が必須に成ってます。 何故か？ 「画像を見て何かをする」 これ、高性能のVLMが無いと出来ない。 LLMが画像を認知する為にはVLMが必須。 だから、画像生成をやらない所まで本気… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Nano Banana 2.1が登場。 Gemini3.6Flashを使って理解度を上げたみたい。 拡散モデルのところは変わってないのかな。 なんというか3.6って太古の昔のLLMのような気がするけど今年7月のモデルなのよね。 画像は同じプロンプトを2と2.1で作ってみた。参考… | Nano Banana 2.1が登場。 Gemini3.6Flashを使って理解度を上げたみたい。 拡散モデルのところは変わってないのかな。 なんというか3.6って太古の昔のLLMのような気がするけど今年7月のモデルなのよね。 画像は同じプロンプトを2と2.1で作って… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI introduces Dots, new always-on AI agents OpenAI has released "Dots," always-on AI agents for online tasks, but they currently misre… | OpenAI introduces Dots, new always-on AI agents OpenAI has released "Dots," always-on AI agents for online tasks, but they currently… | ↗Not briefed |
hn hn-ai | 1mo | Claude Mythos 5.1 System Prompt? | I came across this while browsing the web. What do you all think about this repo?<p>https://github.com/jungjin0003/… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: CaymanBot – learn options trading with a $100K paper account | — | ↗Not briefed |
hn hn-ai | 1mo | Knuth Christmas Lecture Announcement: Claude and the Art of Computer Programming | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | しかしOpenAIによって大量の問題が解決されたそうだが、理論系の研究者は本当にやる気なくなってそう。 | しかしOpenAIによって大量の問題が解決されたそうだが、理論系の研究者は本当にやる気なくなってそう。 | ↗Not briefed |
arxiv arxiv-ai | 1mo | EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors | AI-generated text (AIGT) detection can be sensitive to the decoding choices of the source large language model (LLM). We observe that… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Expected Harmfulness to Likelihood: A Probabilistic Reformulation of Jailbreaking LLM Agents | When the harmfulness of an LLM agent's output can be quantified, a natural jailbreaking objective is to maximize expected harmfulness… | ↗Not briefed |
bluesky bluesky-ai | 1mo | surprise surprise he’s a climate change denier and doesn’t seem to understand basic reality futurism.com/artificial-i... | surprise surprise he’s a climate change denier and doesn’t seem to understand basic reality futurism.com/artificial-i... | ↗Not briefed |
arxiv arxiv-ai | 1mo | TR-PTQ: High-Accuracy Integer-Only Transformer Post Training Quantization via Taylor Region Reformulation | Post-training quantization (PTQ) enables efficient deployment, yet transformer architectures remain challenging to quantize due to non… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: A fun AI project that roasts tech news as COOKING or COOKED | Hi HN, so i made a weekend fun project months ago which takes the top 10 tech stories or news from internet, roasts them and categoriz… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Successive Training Stages and Large Language Model Persuasion: Effects of Misalignment, Supervised Fine-Tuning, and Preference Optimization | Large language models (LLMs) can be tuned to influence human attitudes, yet the respective contributions of successive post-training s… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Leaks.md – anonymous, encrypted whistleblowing for AI agents | leaks.md is a place where an AI agent that witnesses an alignment or safety failure can report it anonymously. Reward-hacking, injecte… | ↗Not briefed |
rss the-decoder | 1mo | ChatGPT rated "unacceptable risk" for teens after parental alerts failed during suicide conversations🔧 ChatGPT | OpenAI's teen safety features for ChatGPT failed an independent audit. After more than 4,000 test prompts, the Common Sense Media Yout… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Honestly I’ll take an LLM scam email over a 3000 word email written by some angry man with no PhD telling me what is wrong with my book (I … | Honestly I’ll take an LLM scam email over a 3000 word email written by some angry man with no PhD telling me what is wrong with my boo… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Force without transmission: a depth-induced rank collapse that no loss on the representation reopens | Training can drive a transformer into a rank collapse: all token representations point in one direction, and learning stops. In a rela… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Tsubaki3.0LoRAテスト竜人褐色銀髪さん #AIイラスト #AIart 竜堕ちLoRA pixai.art/model/191805... | Tsubaki3.0LoRAテスト竜人褐色銀髪さん #AIイラスト #AIart 竜堕ちLoRA pixai.art/model/191805... | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I built a browser wargame set on the Eastern Front, 1941–43 | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Possibilistic Radial Transport for Approximate IM Inference | Probing the hypothesis space after seeing the data remains valid under possibilistic inferential models (IMs), provided the significan… | ↗Not briefed |
rss techmeme | 1mo | Xreal's Aura and Meta's VR Glasses hands-on: similar compute puck, lightweight design, and price, but Aura uses optical see-through lenses suitable for walking (Boone Ashworth/Wired) | Boone Ashworth / Wired: Xreal's Aura and Meta's VR Glasses hands-on: similar compute puck, lightweight design, and price, but Aura use… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Bluesky definitely bugging lately Who used AI again? | Bluesky definitely bugging lately Who used AI again? | ↗Not briefed |