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 |
|---|---|---|---|---|
arxiv arxiv-ai | 1mo | GRACE: Grounded Adversarial Reasoning over Canadian Law | Large language models have shown strong performance across a range of legal tasks, but existing benchmarks rarely evaluate the ability… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: OpensourceDB Tamper-evident,checksum-backed decision and session replay | Audit Trail to make EU AI ACT 12,26 And 50 Compliant Agentic Ai Systems | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: What is up with all the AI doomerism I'm seeing? | I'm constantly seeing posts about how "AI is already better than humans", "programmers are useless" "eve… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I suspect that they are shoving this kind of nonsensical LLMs as these institutions often get funding for their products to be 'AI-powered'… | I suspect that they are shoving this kind of nonsensical LLMs as these institutions often get funding for their products to be 'AI-pow… | ↗Not briefed |
hn hn-ai | 1mo | Lambda MicroEgg | — | ↗Not briefed |
hn hn-ai | 1mo | Misaligned citations and LLM-generated scientific fraud | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Swarm, an offline family tree app for macOS | Hi HN! My wife and I got deep into our family genealogy and wanted to map out all our relatives - but keep the data on our own Macs in… | ↗Not briefed |
hn hn-ai | 1mo | Laya (OS Jev) on Mac M4 CoreML Offline (54 decisions per second) | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I wanted Human Benchmark with global leaderboards, so I built it | Helloooo,<p>I'm a big fan of cognitive games like Human Benchmark and brain games in general.<p>I'm also a big fan of Trackm… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A "radical intentionalist," writes Eryk Salvaggio, treats an AI model as a 'mind' to be interpreted rather than a product that was designed… | A "radical intentionalist," writes Eryk Salvaggio, treats an AI model as a 'mind' to be interpreted rather than a product that was des… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The best explanation of Jev I've found so far 😅 #buildinpublic #jev #typesafeai #aiengineering #LLM #AI | The best explanation of Jev I've found so far 😅 #buildinpublic #jev #typesafeai #aiengineering #LLM #AI | ↗Not briefed |
hn hn-ai | 1mo | Jev compiler – Turn the rules your agent keeps ignoring into gates you can test | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | recording little of that exposure on their balance sheets. First used by Meta on a huge data centre project last autumn, so-called residual… | recording little of that exposure on their balance sheets. First used by Meta on a huge data centre project last autumn, so-called res… | ↗Not briefed |
rss techmeme | 1mo | Vercel, Cloudflare, and others quickly add Jev, as it makes AI tool selection much faster and cheaper; TypeSafe: Jev matches GPT-5.6 and Sonnet 5 workflow evals (Josipa Majic Predin/Forbes)🔧 GPT-5 | Josipa Majic Predin / Forbes: Vercel, Cloudflare, and others quickly add Jev, as it makes AI tool selection much faster and cheaper; T… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Remember when the US military dumped Anthropic because they refused to agree to terms allowing AI to take military action free of human ove… | Remember when the US military dumped Anthropic because they refused to agree to terms allowing AI to take military action free of huma… | ↗Not briefed |
arxiv arxiv-ai | 1mo | STEVE: Stabilizing Textual Gradient-Based Prompt Optimization via Error-Driven Refinement and Regularized Verification🔧 GPT-5 | Textual-gradient methods automate prompt optimization through natural-language feedback, but their iterative updates can be unstable.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | La bande des petits merdeux qui sur les réseaux sociaux se moquent du désarroi des mathématiciens face aux performances des llm, c'est quel… | La bande des petits merdeux qui sur les réseaux sociaux se moquent du désarroi des mathématiciens face aux performances des llm, c'est… | ↗Not briefed |
rss theverge-ai | 1mo | Trump now says he wants to form an ‘AI Force’ | The president posted on Truth Social that he wanted to appoint an "AI czar" to lead a new "AI force." He made the announcement amid gr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI was built on felony. How else do you describe the theft of all that IP? | OpenAI was built on felony. How else do you describe the theft of all that IP? | ↗Not briefed |
hn hn-ai | 1mo | You can cargo build your own codex rust | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | 😁 lol | 😁 lol | ↗Not briefed |
rss the-decoder | 1mo | Tencent's Gander aims to keep talking while it works in the background | Tencent's Gander combines real-time conversation with AI agent capabilities. A "cerebellum" handles the conversation while a swappable… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Lina Khan has pointed out that have robust and well-developed theories of product liability law. What we are missing is *enforcement* www.… | Lina Khan has pointed out that have robust and well-developed theories of product liability law. What we are missing is *enforcement*… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: A 3D Mac window switcher with browser tab support and search | Hey HN. I made this little app, because I missed Compiz but also because I wanted a way to switch between tabs as well as my windows.<… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Real Jev decisions on a simulated robot fleet – $24.57 per million | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Financial Language Models as Applied Artificial Intelligence Systems for News-Based Trading under Market Frictions | Financial language models can transform unstructured firm-specific news into structured decision signals, but financial AI research la… | ↗Not briefed |
bluesky bluesky-ai | 1mo | On the Hugging Face hack, to read the independent post-mortems, the A.I. agents don't come across as Skynet. They come across as terrified …🔧 Hugging Face | On the Hugging Face hack, to read the independent post-mortems, the A.I. agents don't come across as Skynet. They come across as terri… | ↗Not briefed |
bluesky bluesky-ai | 1mo | my post apocalyptic scenario calls for LLM-using raccoons riding on 30-50 wild hogs | my post apocalyptic scenario calls for LLM-using raccoons riding on 30-50 wild hogs | ↗Not briefed |
bluesky bluesky-ai | 1mo | it's essentially a tiny llm that can only answer simple binary or multiple choice questions it'll include its confidence level with those … | it's essentially a tiny llm that can only answer simple binary or multiple choice questions it'll include its confidence level with t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The dispute between the Pentagon and artificial intelligence developer Anthropic that ended in divorce late yesterday had been portrayed as… | The dispute between the Pentagon and artificial intelligence developer Anthropic that ended in divorce late yesterday had been portray… | ↗Not briefed |
arxiv arxiv-ai | 1mo | TEMPER: Temporal Encoder-Masked Probabilistic Ensemble Regressor for Time-Series Forecasting | Probabilistic forecasting requires accurate central predictions and calibrated uncertainty estimates. This paper presents TEMPER, the… | ↗Not briefed |
hn hn-ai | 1mo | ChatGPT now knows what you do on other websites via ad collector🔧 ChatGPT | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Again it’s a very very funny insight into Bluesky that you can make an argument which takes *as a first principle* the nonexistence of soul… | Again it’s a very very funny insight into Bluesky that you can make an argument which takes *as a first principle* the nonexistence of… | ↗Not briefed |
arxiv arxiv-ai | 1mo | When the Agent Becomes the Kernel: A Systematization of Security on the Path to AI-Native Operating Systems | Large language model agents are now privileged principals that take consequential actions: editing code repositories, operating inboxe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | all good. i only pointed out that chuck Schumer's kid got a HUGE job with Anthropic. Which is a whole thing in a very important conversatio… | all good. i only pointed out that chuck Schumer's kid got a HUGE job with Anthropic. Which is a whole thing in a very important conver… | ↗Not briefed |
hn hn-ai | 1mo | Pirate Face Rescues LLM Models from Deletion | — | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: What is one plausible path to 'AI extinction'? | This is being thrown around a lot, including the dreaded "10% chance by the end of this decade". Every conversation I have s… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Distill What You Trust: Reliability-Aware Multi-Teacher On-Policy Distillation | Multi-teacher on-policy distillation allows a student to learn from complementary specialists on its own trajectories. Domain-routed a… | ↗Not briefed |
hn hn-ai | 1mo | Karen Nikoghosyan Is Building AI Operating Layers for Multichannel Commerce | Karen Nikoghosyan works at the crossroads of AI, e-commerce operations, and talent development. As founder and CEO of eSwap Global, he… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PhysAI-Bench: A Benchmark for LLM-Based Agentic Decision-Making in Autonomous UAV-Centric Physical AI🔧 GPT-5 | Recent advances in Physical AI have accelerated the use of foundation models in autonomous systems such as unmanned aerial vehicles (U… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Chesslm – Tiny chess model trained on Stockfish | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Tech companies have been scoffing at model collapse and claiming it’s solved, and it’s not. It’s just now poised to take the entire interne… | Tech companies have been scoffing at model collapse and claiming it’s solved, and it’s not. It’s just now poised to take the entire in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dude is a CEO of an AI company. LOL www.cbc.ca/news/busines... | Dude is a CEO of an AI company. LOL www.cbc.ca/news/busines... | ↗Not briefed |
bluesky bluesky-ai | 1mo | They don’t want to, but if they need to, they’ll figure out how to bring the energy use WAY down. On the other hand, you might want to tak… | They don’t want to, but if they need to, they’ll figure out how to bring the energy use WAY down. On the other hand, you might want t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI cybersecurity risks explode as hackers use Anthropic's #Claude to break into OpenAI. Two weeks after a swarm of AI agents broke out of …🔧 ChatGPT🔧 Hugging Face | AI cybersecurity risks explode as hackers use Anthropic's #Claude to break into OpenAI. Two weeks after a swarm of AI agents broke ou… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I got Claude Code and Codex to argue about my code | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yes. We talk about it at various points but a show about the +s and -s would be a good idea. I primarily use Chatgpt/codex and a few open s…🔧 ChatGPT | Yes. We talk about it at various points but a show about the +s and -s would be a good idea. I primarily use Chatgpt/codex and a few o… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🎥 Privacidade Zero! Suas conversas no ChatGPT ou no Claude podem estar sendo lidas por outras pessoas. OpenAI e Anthropic estariam contra…🔧 ChatGPT | 🎥 Privacidade Zero! Suas conversas no ChatGPT ou no Claude podem estar sendo lidas por outras pessoas. OpenAI e Anthropic estariam c… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Tail-Weight Control and Localized Generalization in Nearly Low-Rank Adversarial Classification | We study norm-constrained linear classification under Eu clidean adversarial perturbations in a Gaussian model with a low-dimen sional… | ↗Not briefed |
rss techmeme | 1mo | A closer look at the upcoming Siri AI-powered home hub, a key pillar of Apple's strategy for the home; sources: Apple started cutting Fitness+ staff (Mark Gurman/Bloomberg) | Mark Gurman / Bloomberg: A closer look at the upcoming Siri AI-powered home hub, a key pillar of Apple's strategy for the home; source… | ↗Not briefed |