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 |
|---|---|---|---|---|
rss wired-ai | 2mo | AI ‘Actor’ Tilly Norwood Told Me That ‘All Lives Matter’ | The virtual character, which is promoting its upcoming movie Misaligned, tries to evade politics by repetitively commenting on the clo… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Amika – Multiplayer cloud workstations for coding agents and humans | --- title: "Show HN: Amika – Multiplayer cloud workstations for coding agents and humans" ---<p>Hi Hacker News!<p>My cofound… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Agentic Societies Need a Social Harness | An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries, on behalf of different principal… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Efficient Robust Learning at the Information-Theoretic Limit | In an important recent work, Blanc (2026) gave an algorithm for robustly learning Boolean concept classes with respect to a fixed dist… | ↗Not briefed |
hn hn-ai | 2mo | Hugging Face is billing OpenAI $100M for hacking it🔧 Hugging Face | — | ↗Not briefed |
rss bing-news-ai | 2mo | Artificial intelligence industry leaders are calling for safety guardrails | UNR professor says AI guardrails will build public trust and build a system of accountability ... | ↗Not briefed |
rss bing-news-ai | 2mo | Artificial intelligence industry leaders are calling for safety guardrails | UNR professor says AI guardrails will build public trust and build a system of accountability ... | ↗Not briefed |
arxiv arxiv-ai | 2mo | ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents | We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agent… | ↗Not briefed |
arxiv arxiv-ai | 2mo | PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control | Interactive control for video generation is moving from coarse prompts toward fine-grained, physically meaningful manipulation of dyna… | ↗Not briefed |
bluesky bluesky-ai | 2mo | (4/6) (1) requiring large AI firms to report critical safety incidents in a timely manner (2) requiring independent safety reviews before … | (4/6) (1) requiring large AI firms to report critical safety incidents in a timely manner (2) requiring independent safety reviews be… | ↗Not briefed |
arxiv arxiv-ai | 2mo | When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control | Large language models can produce fluent answers when their factual support is weak. This paper introduces Chain-of-Self-Questioning (… | ↗Not briefed |
rss the-decoder | 2mo | AI labs have a data trust problem that their policies haven't solved | OpenAI and Anthropic tell corporate customers their data won't be used for training. But when Anthropic said it would store usage logs… | ↗Not briefed |
arxiv arxiv-ai | 2mo | What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity | Pruning can reduce the deployment cost of large language models (LLMs), but its impact on context-grounded tool calling remains poorly… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Well if you want to become less confused about the Anthropic thing definitely don't call the number in that email because I just had the mo… | Well if you want to become less confused about the Anthropic thing definitely don't call the number in that email because I just had t… | ↗Not briefed |
hn hn-ai | 2mo | Public Data for Health at the OpenAI Foundation | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Pede uma ferramenta do Google Abraji, chama um rel gov da Meta, um policy manager da OpenAI... | Pede uma ferramenta do Google Abraji, chama um rel gov da Meta, um policy manager da OpenAI... | ↗Not briefed |
arxiv arxiv-ai | 2mo | LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs | Neural audio codecs are a key component in speech language modeling. However, their high frame rates lead to long sequence lengths, in… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Ben Jennings on tech leaders’ outcry over the dangers of AI – cartoon | Ben Jennings on tech leaders’ outcry over the dangers of AI – cartoon | ↗Not briefed |
bluesky bluesky-ai | 2mo | Oh god, the $600 beer money I got over two years of fact-checking LLM responses, what a conflict of interest | Oh god, the $600 beer money I got over two years of fact-checking LLM responses, what a conflict of interest | ↗Not briefed |
bluesky bluesky-ai | 2mo | Latest @theguardian.com cartoon www.theguardian.com/commentisfre... | Latest @theguardian.com cartoon www.theguardian.com/commentisfre... | ↗Not briefed |
rss techcrunch-ai | 2mo | AI Agents now have a place to snitch | The AI Contact Hotline is designed to be a discreet place where agents that have witnessed misbehavior can tip off authorities. | ↗Not briefed |
arxiv arxiv-ai | 2mo | ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation | Uncertainty estimation for Vision-Language-Navigation (VLN) models is a critical task since it can help identify ambiguous and unrelia… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Regularized Least Squares Training of Quadratic Neural Networks with Applications to System Identification | This paper proposes a least squares approach for the training of quadratic neural networks with regularization. The proposed methodolo… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I used to do gig work with Data Annotation for cash on the side, checking LLM outputs for errors, doing different kinds of tests. I saw the… | I used to do gig work with Data Annotation for cash on the side, checking LLM outputs for errors, doing different kinds of tests. I sa… | ↗Not briefed |
hn hn-ai | 2mo | Gemini 3.8 Live and 3.8 Live Extended Thinking | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | juli an llm parses the above post correctly and i have faith that you can do so too. try again | juli an llm parses the above post correctly and i have faith that you can do so too. try again | ↗Not briefed |
arxiv arxiv-ai | 2mo | Verifiable Social Reasoning for LLM Assistants | LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation settings remains ch… | ↗Not briefed |
hn hn-ai | 2mo | Why I'm still bearish on LLMs after navier-stokes | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Some developers are using the Claude Code harness to access cheaper non-Anthropic models, including GPT-5.6 Sol, via proxies and services l…🔧 OpenRouter🔧 GPT-5 | Some developers are using the Claude Code harness to access cheaper non-Anthropic models, including GPT-5.6 Sol, via proxies and servi… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Obviously this is going to be pilloried on Bluesky, but if you work in the PMO are you calling this headline a win, or is it an awkward dis… | Obviously this is going to be pilloried on Bluesky, but if you work in the PMO are you calling this headline a win, or is it an awkwar… | ↗Not briefed |
rss techmeme | 2mo | Some developers are using the Claude Code harness to access cheaper non-Anthropic models, including GPT-5.6 Sol, via proxies and services like OpenRouter (Alix Coutures/The Information)🔧 OpenRouter🔧 GPT-5 | Alix Coutures / The Information: Some developers are using the Claude Code harness to access cheaper non-Anthropic models, including G… | ↗Not briefed |
bluesky bluesky-ai | 2mo | At the Canada Investment Summit, I got some reaction from OpenAI executive... and the former UK Chancellor of the Exchequer who hired Mark … | At the Canada Investment Summit, I got some reaction from OpenAI executive... and the former UK Chancellor of the Exchequer who hired… | ↗Not briefed |
hn hn-ai | 2mo | Lyric – Windows dictation with a guard against LLM hallucinations | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | 50% of the LLM must be in French | 50% of the LLM must be in French | ↗Not briefed |
bluesky bluesky-ai | 2mo | So OpenAi wants to delay their IPO. Safety concerns about Frontier AI are real, but they can also provide a publicly acceptable rationale… | So OpenAi wants to delay their IPO. Safety concerns about Frontier AI are real, but they can also provide a publicly acceptable rati… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Self-adjusting vLLM at production scale🔧 vLLM | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | FreqSpaNet: Frequency and Spatial Learning of SFPF for Physical Layer Hardware Integrity Detection | Unauthorized hardware replacement can preserve a wireless device's logical identity while altering its physical implementation, posing… | ↗Not briefed |
bluesky bluesky-ai | 2mo | also i feel like we have lost track of the whole "it is impossible to store all the information that an LLM is trained on inside the LLM, i… | also i feel like we have lost track of the whole "it is impossible to store all the information that an LLM is trained on inside the L… | ↗Not briefed |
arxiv arxiv-ai | 2mo | LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence | We introduce LimiX-2, a new model in the LimiX family, developed through model and data scaling guided by our previously established s… | ↗Not briefed |
hn hn-ai | 2mo | Rightsholders Can't Use OpenAI and Anthropic to Dismantle Meta's Seeding Defense | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Free tool that checks if your Etsy/Shopify listing is EU-ready | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | LLM Interpretability, a famously solved subfield | LLM Interpretability, a famously solved subfield | ↗Not briefed |
hn hn-ai | 2mo | Qector-mega 2.2.1 – Sub-ms QEC decoding on a 2-core CPU with zero GPU needs | — | ↗Not briefed |
hn hn-ai | 2mo | Ask HN: Harness for Learning – Thoughts? | I've been thinking about a harness for the learning use case. Current harness objective is to basically take a task and drive it… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Maybe, but the average understanding of an LLM by a Bsky user seems to be stuck in 2024. I still people posting the "strawberry" bullshit … | Maybe, but the average understanding of an LLM by a Bsky user seems to be stuck in 2024. I still people posting the "strawberry" bull… | ↗Not briefed |
rss techmeme | 2mo | Thatch, a healthcare plan marketplace where employers set a fixed budget that employees can use for individual plans, raised a $108M Series C at a $1B valuation (Marina Temkin/TechCrunch) | Marina Temkin / TechCrunch: Thatch, a healthcare plan marketplace where employers set a fixed budget that employees can use for indivi… | ↗Not briefed |
bluesky bluesky-ai | 2mo | OPENAI EXECUTIVE GEORGE OSBORNE CALLED CARNEY'S AI ADOPTION STRATEGY ONE OF THE CLEAREST AMONG WESTERN GOVERNMENTS, SIGNALING CANADA COULD … | OPENAI EXECUTIVE GEORGE OSBORNE CALLED CARNEY'S AI ADOPTION STRATEGY ONE OF THE CLEAREST AMONG WESTERN GOVERNMENTS, SIGNALING CANADA C… | ↗Not briefed |
hn hn-ai | 2mo | PyMalbolge: Malbolge Interpreter in Python | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Bridging the Gap Between Homogeneous and Heterogeneous Asynchronous Optimization Is Surprisingly Difficult | Modern large-scale machine learning tasks often require multiple workers, devices, CPUs, or GPUs to compute stochastic gradients in pa… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents | GUI agents execute long-horizon tasks on dynamic graphical user interfaces, where pop-ups, delayed loads, and relocated widgets routin… | ↗Not briefed |