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 | 9d | Orpheus.el: Emacs Transient Music Player (SSH, MPD, Local, YouTube, etc.) | — | ↗Not briefed |
arxiv arxiv-ai | 9d | RMIT-ADM+S at TREC RAG 2026: Research Agents, Retriever Ablations, and Open-Weight Replication | We describe the RMIT-ADM+S team's participation in the TREC 2026 Retrieval-Augmented Generation (RAG) track using the ClimbMix corpus.… | ↗Not briefed |
bluesky bluesky-ai | 9d | 香港で話題に出た日本のこと 1. VIVANT 結構見てる人が多く、見ている人は一人で描く漫画が素晴らしいのになんで愛国ポルノを作ってしまうのかと笑っている 話題といえば「高市」というフレーズは何度も出てきた。これは実害を感じているからだから仕方がない 2. 日本の国産AI… | 香港で話題に出た日本のこと 1. VIVANT 結構見てる人が多く、見ている人は一人で描く漫画が素晴らしいのになんで愛国ポルノを作ってしまうのかと笑っている 話題といえば「高市」というフレーズは何度も出てきた。これは実害を感じているからだから仕方がない 2. 日本… | ↗Not briefed |
arxiv arxiv-ai | 9d | PhysCapBench: A Cross-Domain Benchmark for Physically Grounded Video Captioning | Physical AI aims to learn transferable physical knowledge from real-world experience to understand, predict, and act in the physical w… | ↗Not briefed |
arxiv arxiv-ai | 9d | BlinkMimic: Training a Humanoid Motion Tracker in Minutes on a Single GPU | Whole-body motion tracking lets a humanoid robot follow reference motions, but training a tracking policy for a library of motions has… | ↗Not briefed |
rss techmeme | 9d | A growing number of young victims whose photos were sexualized using Grok are seeking to hold xAI legally accountable for allowing users to generate such images (Alexandra S. Levine/Bloomberg) | Alexandra S. Levine / Bloomberg: A growing number of young victims whose photos were sexualized using Grok are seeking to hold xAI leg… | ↗Not briefed |
hn hn-ai | 9d | Why Did Palantir CEO Alex Karp Buy a Swedish Forest? | — | ↗Not briefed |
arxiv arxiv-ai | 9d | Small Language Models for Inference-Capability Routing: Effects of Model Capacity and Prompt Optimization on Downstream Quality and Efficiency | Optional reasoning, retrieval, web search, and tool use can improve large language model responses but incur unnecessary computation w… | ↗Not briefed |
bluesky bluesky-ai | 9d | Krožno referenčno digitalno (zeleno malo manj) gospodarstvo. Predvsem krožno. www.reuters.com/world/china/... | Krožno referenčno digitalno (zeleno malo manj) gospodarstvo. Predvsem krožno. www.reuters.com/world/china/... | ↗Not briefed |
hn hn-ai | 9d | Show HN: Stop Safari auto zoom on input fields | A very simple Safari extension for iOS that disables the automatic zooming in on Safari when focusing an input field with less than 16… | ↗Not briefed |
bluesky bluesky-ai | 9d | Bref tout le monde s’en fout car là c’est de l’IA pour la recherche, pour la science ; pas d’effet d’annonce façon OpenAI ou Anthropic… E… | Bref tout le monde s’en fout car là c’est de l’IA pour la recherche, pour la science ; pas d’effet d’annonce façon OpenAI ou Anthropic… | ↗Not briefed |
hn hn-ai | 9d | Why Every AI Model Will Ship as a Torrent | — | ↗Not briefed |
arxiv arxiv-ai | 9d | Experience-Handoff Reinforcement Learning for Recursive Self-Improvement | It is common to train Large Language Models (LLMs) for agentic tasks with Reinforcement Learning (RL) from scalar rewards. However, th… | ↗Not briefed |
hn hn-ai | 9d | My memory isn't getting worse, it's just using exponential decay | — | ↗Not briefed |
bluesky bluesky-ai | 9d | Zero-Dependency SDK Fallback Routing vs LiteLLM & OpenRouter JSON How to eliminate supply-chain risk and cut blended LLM API bills by 68% …🔧 OpenRouter | Zero-Dependency SDK Fallback Routing vs LiteLLM & OpenRouter JSON How to eliminate supply-chain risk and cut blended LLM API bills by… | ↗Not briefed |
rss techmeme | 9d | Munich-based Agile Robots, which develops humanoids, mobile robots, robotic arms, and control software, raised $500M led by SoftBank at an $8.5B valuation (Marilen Martin/Bloomberg) | Marilen Martin / Bloomberg: Munich-based Agile Robots, which develops humanoids, mobile robots, robotic arms, and control software, ra… | ↗Not briefed |
arxiv arxiv-ai | 9d | Saidzoda-Kazakh-EmbeddingGemma-300M: How Far One Epoch of Generated Questions Takes a Small Kazakh Embedder🔧 Hugging Face | EmbeddingGemma-300M is a small open embedder, and used as released it finds the right passage first for 60.8% of the Kazakh Belebele r… | ↗Not briefed |
bluesky bluesky-ai | 9d | Neben Anthropic und Apple CEO wahrscheinlich der beste Tech-Job, den man haben kann und das alles wegen Feelings verlieren | Neben Anthropic und Apple CEO wahrscheinlich der beste Tech-Job, den man haben kann und das alles wegen Feelings verlieren | ↗Not briefed |
arxiv arxiv-ai | 9d | TransXplainRRG: A Radiologist-Evaluated LVLM for Chest Radiograph Report Generation | Recent advancements in Large Vision-Language Models (LVLM) enhance X-ray report generation, but often neglect clinical context, overlo… | ↗Not briefed |
bluesky bluesky-ai | 9d | 还打算买几只anthropic的ipo,只有我参与了才能触发AI泡沫的破灭,人类应该可以感谢我 | 还打算买几只anthropic的ipo,只有我参与了才能触发AI泡沫的破灭,人类应该可以感谢我 | ↗Not briefed |
bluesky bluesky-ai | 9d | Man Jensen Huang might be going to prison not just getting ousted This is going to be a MAJOR rug pull from openAI and Anthropic, even Elo… | Man Jensen Huang might be going to prison not just getting ousted This is going to be a MAJOR rug pull from openAI and Anthropic, eve… | ↗Not briefed |
hn hn-ai | 9d | Show HN: OpenPocket – iOS app for running Qwen3.5, LFM and other LLMs offline | I built OpenPocket, a completely free iOS app for running open-weight LLMs locally on iPhones and iPads using Apple's MLX framewo… | ↗Not briefed |
rss techmeme | 9d | Radical Numerics unveils Omnii Defense, an edge AI model it says can identify natural and AI-engineered pathogens without an internet connection (Emily Mullin/Wired) | Emily Mullin / Wired: Radical Numerics unveils Omnii Defense, an edge AI model it says can identify natural and AI-engineered pathogen… | ↗Not briefed |
bluesky bluesky-ai | 9d | ngl if they name a model Celestial Pivot I'm getting a deepseek subscription immediately | ngl if they name a model Celestial Pivot I'm getting a deepseek subscription immediately | ↗Not briefed |
bluesky bluesky-ai | 9d | also i am pretty sure if you fed an LLM my bluesky posting history and asked it to guess what game i was grousing about it would guess corr… | also i am pretty sure if you fed an LLM my bluesky posting history and asked it to guess what game i was grousing about it would guess… | ↗Not briefed |
arxiv arxiv-ai | 9d | Beyond Compressing Everything: Contrastive Memory Learning for Selective Context Compression | As large language models (LLMs) advance toward increasingly long inputs, soft context compression offers a promising approach to effic… | ↗Not briefed |
bluesky bluesky-ai | 9d | Yooo this is more scandalous than the Honeyplot. Disappearing sales and financial statements!?!?! The LLM bubble is cooked!!! | Yooo this is more scandalous than the Honeyplot. Disappearing sales and financial statements!?!?! The LLM bubble is cooked!!! | ↗Not briefed |
hn hn-ai | 9d | I made a CLI focus music player with lo-fi, Pomodoro timer, and Spotify support | — | ↗Not briefed |
hn hn-ai | 9d | Worse Together: How Performance Breaks Down in Multi-User Multi-Agent Teams | — | ↗Not briefed |
bluesky bluesky-ai | 9d | DeepSeek V4.1 Flash vs GPT-6 Luna Pro: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the live 2026 token m… | DeepSeek V4.1 Flash vs GPT-6 Luna Pro: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the live 2026 to… | ↗Not briefed |
arxiv arxiv-ai | 9d | Evidence-Governed Adoption of Self-Generated Updates in Autonomous AI Agents | Autonomous agents modify prompts, memories, tools, workflows, and policies using interaction histories. Yet a plausible modification i… | ↗Not briefed |
hn hn-ai | 9d | LeapQuant: Efficient Linear Attention with Accurate Recurrent State Quantization | — | ↗Not briefed |
arxiv arxiv-ai | 9d | Recursive State-Augmented Language Models: Persistent Latent-State Adaptation of Pretrained Language Models | Language models rely on finite token contexts, limiting access to earlier interactions. We introduce Recursive State-Augmented Languag… | ↗Not briefed |
hn hn-ai | 9d | Ask HN: Are clean Git commits still worth it? | Git is obvioulsy good for sharing/storing code or removing changes you/AI just made. But do we still need to make small, sel… | ↗Not briefed |
bluesky bluesky-ai | 9d | ローカルLLM(道の駅に売っている) | ローカルLLM(道の駅に売っている) | ↗Not briefed |
hn hn-ai | 9d | 258 Nvidia headlines: 6% ethics-negative before the scandal, 82% after | — | ↗Not briefed |
bluesky bluesky-ai | 9d | I wrote about this on the other account, but my struggle to kick LLM usage once I tried it, even for a little bit just to understand it, wa… | I wrote about this on the other account, but my struggle to kick LLM usage once I tried it, even for a little bit just to understand i… | ↗Not briefed |
bluesky bluesky-ai | 9d | Googleが意思決定AIモデルの「Gemini Decisions 1」をリリース https://gigazine.net/news/20261112-gemini-decisions-1/ | Googleが意思決定AIモデルの「Gemini Decisions 1」をリリース https://gigazine.net/news/20261112-gemini-decisions-1/ | ↗Not briefed |
arxiv arxiv-ai | 9d | Bait: Mitigating Harmful Agentic Reinforcement Learning via Probing Reward Function | We expose a safety vulnerability on nascent agentic reinforcement-learning-as-a-service. Specifically, with a cost as low as \$20, we… | ↗Not briefed |
arxiv arxiv-ai | 9d | ReliaGuard: Reconstruction-Aware Watermarking for Source Attribution and Blind Edit Localization | Generative image editing facilitates content creation but complicates source attribution and edit verification. Existing watermarking… | ↗Not briefed |
bluesky bluesky-ai | 9d | like fine don’t give a fucking award to some dumbshit game because they used an llm, idgaf, please find a better catchphrase while i try to… | like fine don’t give a fucking award to some dumbshit game because they used an llm, idgaf, please find a better catchphrase while i t… | ↗Not briefed |
hn hn-ai | 9d | Show HN: JNum – Pure Java 25 NumPy/BLAS via Project Panama FFM and Vector API | — | ↗Not briefed |
hn hn-ai | 9d | AI is getting smarter. Are we thinking better? | — | ↗Not briefed |
bluesky bluesky-ai | 9d | Open source software is a pretty good space to see what it could look like. Though, at present, that 'common' intellectual property just ge… | Open source software is a pretty good space to see what it could look like. Though, at present, that 'common' intellectual property ju… | ↗Not briefed |
hn hn-ai | 9d | Show HN: Minimal and self-hosted website monitoring | Hey everyone!<p>I run multiple VPSes for various side projects. One of my servers hosts all of my unpopular/low-traffic apps, and… | ↗Not briefed |
hn hn-ai | 9d | Ask HN: Is GitHub Downloads Down? | Facing issues downloading release artifacts since today morning. Status pages show everything is okay. A 68MB release artifact took me… | ↗Not briefed |
arxiv arxiv-ai | 9d | MutAtlas: A PDB-Wide Energy-Guided Atlas of Protein Mutation Effects | Protein mutation effect prediction is fundamental to protein engineering and disease variant interpretation, yet experimentally measur… | ↗Not briefed |
arxiv arxiv-ai | 9d | IsharaBART: An efficient BART-Based Framework for Low-Resource Bangla Sign Language Translation | Millions of Bangla Sign Language (BdSL) users lack reliable automatic translation tools. Existing systems are limited by data scarcity… | ↗Not briefed |
hn hn-ai | 9d | What benefit does a San Francisco HQ offer if you are not a frontier AI lab? | I am struggling to see the benefit of basing your startup in SF if you are not a frontier AI lab? It seems to me that the talent quali… | ↗Not briefed |
hn hn-ai | 9d | Redis Got Ported to Rust | — | ↗Not briefed |