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 | JEV-as-a-Judge: Accept When Confident, Escalate When Unsure | LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence reliability become critical at scale. We stu… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Design 3D printable containers that fit any drawer perfectly | I built Container CAD to design containers that fit exactly into any drawer easier.<p>For me, it solved:<p>1. Wasting hours on CAD des… | ↗Not briefed |
bluesky bluesky-ai | 1mo | please don’t post a huge wall of clearly llm written text with multiple factual inaccuracies at another developer you ostensibly want to wo… | please don’t post a huge wall of clearly llm written text with multiple factual inaccuracies at another developer you ostensibly want… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Should be doable with a regular web scraping-like bot. I’ve defeated more onerous protections before (even pre-LLM) | Should be doable with a regular web scraping-like bot. I’ve defeated more onerous protections before (even pre-LLM) | ↗Not briefed |
arxiv arxiv-ai | 1mo | Topology-Stratified Materials Discovery with A Flow-Based Generative Model | Accurate generation of crystal structures is the foundation to the discovery of high-performance materials for extreme-environment app… | ↗Not briefed |
hn hn-ai | 1mo | GAN-based colorization of SAR imagery into optical-style RGB | — | ↗Not briefed |
rss techcrunch-ai | 1mo | Five AI safety sessions every founder should have on their TechCrunch Disrupt 2026 agenda | At TechCrunch Disrupt 2026, five sessions across the AI Stage and Real World AI Stage cover AI safety, featuring leaders from Anthropi… | ↗Not briefed |
rss techcrunch-ai | 1mo | AstroForge is putting AI in command of its next spacecraft | Autonomy-1 will have a small, transformer-based AI model taking charge of a space probe. | ↗Not briefed |
bluesky bluesky-ai | 1mo | It's long past time to let go all US music writers and replace them with a much more reliable LLM trained on the UK press and fanzines. | It's long past time to let go all US music writers and replace them with a much more reliable LLM trained on the UK press and fanzines. | ↗Not briefed |
arxiv arxiv-ai | 1mo | A retrospective analysis on the use of LLMs to study infant syntax learning | Large language models (LLMs) have increasingly been used to investigate how children acquire syntax at an early stage of development.… | ↗Not briefed |
rss bing-news-ai | 1mo | Is it ‘killer robots’ or ‘super intelligence’? At the UN, leaders see both | World leaders are painting quite different views of the future of artificial intelligence at the United Nations, with colorful phrases… | ↗Not briefed |
rss bing-news-ai | 1mo | Is it ‘killer robots’ or ‘super intelligence’? At the UN, leaders see both | World leaders are painting quite different views of the future of artificial intelligence at the United Nations, with colorful phrases… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Notes on Fourier-Bessel wavelets | These notes develop the mathematical foundations and construction of a Fourier-Bessel wavelet family inspired by the disk harmonics of… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Is AI use not allowed in projects which is shared here to show? | Dear members, I saw someone show HN project got flagged and people were commenting it is probably because it looked AI slop. I'm… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Transcribe, Translate, and Optimize: Joint Reward Learning for Speech Translation | In LLM-based speech translation, transcription-based chain-of-thought (CoT) suffers from a mismatch between reference transcripts used… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Whami is a daily trivia game that writes its own questions every night | For my Masters final project I had WHAMI built by offshore developers and it took 2 months. I recently rebuilt it myself with Claude… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: CircuitPilot – We built a physics engine to auto-route PCBs using LLMs | Hey HN,<p>We wanted to see if we could build an AI agent that actually designs real, manufacturable PCBs from natural language, rather… | ↗Not briefed |
arxiv arxiv-ai | 1mo | REFLEX with Jev for Efficient Selective Control in LLM Agents | LLM agents often use generative models for bounded decisions, raising the question of when these decisions can be handled more efficie… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Another sad case of a human who is dismissive of LLMs capabilities getting something wrong bc they didn’t check it with an LLM before posti… | Another sad case of a human who is dismissive of LLMs capabilities getting something wrong bc they didn’t check it with an LLM before… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 見かけたので。 えーと。 「OpenAIのAIを訓練している人が、AIの訓練にAIを使って解雇された」 でいいのかな。 2026年9月22日 404Media 疲れて頭が働いてないな…。 とんちか早口言葉かな、ってなってる。 www.404media.co/peopl… | 見かけたので。 えーと。 「OpenAIのAIを訓練している人が、AIの訓練にAIを使って解雇された」 でいいのかな。 2026年9月22日 404Media 疲れて頭が働いてないな…。 とんちか早口言葉かな、ってなってる。 www.404media.co/… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Semiotics-Aware Framework for Evaluating Fidelity and Coverage in Natural Language Generation | When two texts describe the same expression, standard metrics based on lexical overlap or whole-text similarity may fail to detect mea… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Piary – Rethinking the human-agent interface beyond chat🔧 ChatGPT | Hello HN! We are Nemo and Baran. We are interaction designers based in Zurich and have been exploring what future agentic interfaces s… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm gonna locally run an LLM of your mom | I'm gonna locally run an LLM of your mom | ↗Not briefed |
rss techmeme | 1mo | Snorkel AI, which offers an "agentic data development platform" with humans and AI agents to create and vet data, raised $350M at a $3.5B valuation (Krystal Hu/Reuters) | Krystal Hu / Reuters: Snorkel AI, which offers an “agentic data development platform” with humans and AI agents to create and vet data… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump pitches ‘super intelligence’ as new name for AI | President Donald Trump sought to rebrand artificial intelligence as “super intelligence” Tuesday in remarks to the UN General Assembly… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump pitches ‘super intelligence’ as new name for AI | President Donald Trump sought to rebrand artificial intelligence as “super intelligence” Tuesday in remarks to the UN General Assembly… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump pitches ‘super intelligence’ as new name for AI | President Donald Trump sought to rebrand artificial intelligence as “super intelligence” Tuesday in remarks to the UN General Assembly… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They are at least consistent though. Every bad thing is the fault of people without power, while the powerful get a pass. Should the gove… | They are at least consistent though. Every bad thing is the fault of people without power, while the powerful get a pass. Should the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Have been thinking a LOT about this outright theft. Context is news scraping but AI has stolen from so many authors, too. And in Anthropic … | Have been thinking a LOT about this outright theft. Context is news scraping but AI has stolen from so many authors, too. And in Anthr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I have in fact been repeatedly advised that model collapse isn't real, the Hapsburg syndrome isn't real and that me thinking that training … | I have in fact been repeatedly advised that model collapse isn't real, the Hapsburg syndrome isn't real and that me thinking that trai… | ↗Not briefed |
hn hn-ai | 1mo | OpenAI is about to eat Jev's lunch – Arcturus Labs | — | ↗Not briefed |
hn hn-ai | 1mo | Jev – a curation of Jev demos on X, tools, skills, and integrations | — | ↗Not briefed |
hn hn-ai | 1mo | Actively Inducing Side-Channel Leakage Using Electromagnetic Injection | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Do Vision Model See Like the Brain? A Comparison Across EEG Encoding Model | Convolutional neural networks (CNNs) and vision transformers are both used to model the human visual system, but whether the two archi… | ↗Not briefed |
hn hn-ai | 1mo | Watch-Wearing as a Marker of Conscientiousness | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | um excuse me I was assured that including the output of an LLM into the training data like a butt sewn to a mouth was in fact a good idea, … | um excuse me I was assured that including the output of an LLM into the training data like a butt sewn to a mouth was in fact a good i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Jaw-dropping documents from our lawsuit against Microsoft and OpenAI reveal how tech leaders talk when they think no one is listening: We m… | Jaw-dropping documents from our lawsuit against Microsoft and OpenAI reveal how tech leaders talk when they think no one is listening:… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ON a completely unrelated note: "Anthropic has quietly set up a laboratory to do physical biology work as it pushes its AI ambitions into … | ON a completely unrelated note: "Anthropic has quietly set up a laboratory to do physical biology work as it pushes its AI ambitions… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Gap-Free Streaming PCA Beyond Rank-One Updates: Near-Optimal Rates and Applications to Differential Privacy | Streaming principal component analysis (PCA) seeks to recover a leading spectral subspace in a single pass over a data stream. We give… | ↗Not briefed |
arxiv arxiv-ai | 1mo | The Ethics of Artificial Intelligence in Military Operations | Deep learning systems now mediate military decisions to use force, yet their internal logic resists inspection, their evaluation pract… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Teams default-route to the strongest model, paying premium fees on routine work. The split is the discipline: premium 10-15%, workhorse 50-… | Teams default-route to the strongest model, paying premium fees on routine work. The split is the discipline: premium 10-15%, workhors… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Numban reports scammers callback number rather than spoofed caller ID | I get a LOT of scam calls every week and cannot reliably block them because they spoof their caller ID. But they CANNOT spoof the call… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Deep Generative Crystal Structure Prediction: A Benchmark Study and a Controlled Test of Prototype Dependence🔧 Consensus | Deep generative models are widely reported to enable de novo crystal structure prediction (CSP), but their capability has not been mea… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Brayer – Create product mockups locally in the browser | Hey HN, I built Brayer to make mockups for my other SaaS projects. I couldn't get existing tools to do what I wanted, so I starte… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is because AI eschatology assumes a very specific model of intelligence, a utility-maximizer, and it turns out that’s not the shape ou… | this is because AI eschatology assumes a very specific model of intelligence, a utility-maximizer, and it turns out that’s not the sha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The quotes in this are totally bonkers. We sued OpenAI. Now, court documents are revealing what AI execs apparently knew and when—about the… | The quotes in this are totally bonkers. We sued OpenAI. Now, court documents are revealing what AI execs apparently knew and when—abou… | ↗Not briefed |
bluesky bluesky-ai | 1mo | NEW: When OpenAI claimed to solve one of the toughest problems in math earlier this month it promised the results would "benefit all of hum… | NEW: When OpenAI claimed to solve one of the toughest problems in math earlier this month it promised the results would "benefit all o… | ↗Not briefed |
rss techcrunch-ai | 1mo | TechCrunch Disrupt 2026: Aaron Edsinger brings Hello Robot’s Stretch 4 to life onstage | Hello Robot CEO and co-founder Aaron Edsinger will bring Stretch 4 for a live demo on the Real World AI Stage at TechCrunch Disrupt 20… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: DrivingBench – Frontier LLMs Driving a Real Toyota Corolla🔧 Grok 4 | Hi all, wanted to share a cool project me and two buddies have been working on. Essentially we took a Toyota Corolla, installed a Comm… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Radiomics-Conditioned Modulation of RenalCLIP Features for Clear Cell Renal Cell Carcinoma Classification | Radiomics provides quantitative descriptions of tumour appearance that may complement disease-specific foundation models in small labe… | ↗Not briefed |