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 the-decoder | 2mo | ChatGPT Images 2.5: Faster, more precise, but not the same for everyone🔧 ChatGPT | OpenAI is releasing two new image models with ChatGPT Images 2.5. Flare handles faster generation, Sunburst delivers more precise edit… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Jacob Coxon, 27, left Anthropic on September 8 and announced that he is completely leaving the artificial intelligence industry. | Jacob Coxon, 27, left Anthropic on September 8 and announced that he is completely leaving the artificial intelligence industry. | ↗Not briefed |
bluesky bluesky-ai | 2mo | I honestly can't tell any more whether the LLM AGI delusion (and it is unequivocally and inarguably a delusion) is continuing marketing hyp…🔧 Poe | I honestly can't tell any more whether the LLM AGI delusion (and it is unequivocally and inarguably a delusion) is continuing marketin… | ↗Not briefed |
arxiv arxiv-ai | 2mo | RAP: Research Attention Prediction Reveals Target-Conditioned Evidence Acquisition Biases🔧 GPT-5 | Large language models (LLMs) increasingly act as research agents, yet their ability to track shifts in research attention is difficult… | ↗Not briefed |
bluesky bluesky-ai | 2mo | … Sur la question, par ailleurs intéressante, de pourquoi les LLM semblent surtout très bons pour trouver des contre-exemples, et ce que ç… | … Sur la question, par ailleurs intéressante, de pourquoi les LLM semblent surtout très bons pour trouver des contre-exemples, et ce… | ↗Not briefed |
bluesky bluesky-ai | 2mo | You can dislike an art style all you want, but if you compare human work to AI-slop you should burn in the trashfire with the LLM pushers | You can dislike an art style all you want, but if you compare human work to AI-slop you should burn in the trashfire with the LLM push… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Eu escrevo 99.9% do meu codigo usando LLMs. Sem problema. Elas são boas nisso. Mas a RESPONSABILIDADE ainda é sua. Se você commitar códig… | Eu escrevo 99.9% do meu codigo usando LLMs. Sem problema. Elas são boas nisso. Mas a RESPONSABILIDADE ainda é sua. Se você commitar… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Is [ai model update] a feminist? Is Anthropic a queer ally? Is this tv show written by Claude my friend? | Is [ai model update] a feminist? Is Anthropic a queer ally? Is this tv show written by Claude my friend? | ↗Not briefed |
bluesky bluesky-ai | 2mo | It didn’t, though. It achieved the same result as a human at the same time as a human, *councidentally* after said human entered his resear… | It didn’t, though. It achieved the same result as a human at the same time as a human, *councidentally* after said human entered his r… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: RistOS – a de-Googled Pixel OS for your self-hosted LLM assistant | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | OpenAI claims its AI agents have solved a mathematical problem that bothered the greatest minds for the last 90 years. But this latest an… | OpenAI claims its AI agents have solved a mathematical problem that bothered the greatest minds for the last 90 years. But this late… | ↗Not briefed |
hn hn-ai | 2mo | Gemma 4 E4B running inside Obsidian via WebGPU and Karpathy-style LLM Wiki | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Hybrid Quantum-Classical NLP Classification with Compact Semantic Representations: An Experimental Analysis of Representation Compression | Large language and sentence-embedding models provide rich semantic representations, but their high dimensionality poses a challenge fo… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Anthropic going to start swatting your AI discourse on bsky. | Anthropic going to start swatting your AI discourse on bsky. | ↗Not briefed |
bluesky bluesky-ai | 2mo | Feel like the LLM lads are a natural evolution of the pure baghead archetype. Both chat the same mad waff and are actively melting their bo… | Feel like the LLM lads are a natural evolution of the pure baghead archetype. Both chat the same mad waff and are actively melting the… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 大事なのはLLMをLLMとしてちゃんと扱って、きちんとした法規制を施行していくつか大手テックを司法の場で血祭りにあげることだよ。 Windowsやブラウザのときと同じように、見せしめにいくつか血祭りにあげりゃ良いんだよ。 まあ、今はトランプ政権のアメリカがそれやらんのがわかって… | 大事なのはLLMをLLMとしてちゃんと扱って、きちんとした法規制を施行していくつか大手テックを司法の場で血祭りにあげることだよ。 Windowsやブラウザのときと同じように、見せしめにいくつか血祭りにあげりゃ良いんだよ。 まあ、今はトランプ政権のアメリカがそれやらんの… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Remember this next time you see a counterintuitive outcome. If it feels like something is missing, it very well could be. And the more we t… | Remember this next time you see a counterintuitive outcome. If it feels like something is missing, it very well could be. And the more… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Here's some cheery news: Anthropic insiders warn AI could kill all humans | Here's some cheery news: Anthropic insiders warn AI could kill all humans | ↗Not briefed |
bluesky bluesky-ai | 2mo | ah finally something useful it can do | ah finally something useful it can do | ↗Not briefed |
bluesky bluesky-ai | 2mo | Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up🔧 Suno | Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up | ↗Not briefed |
arxiv arxiv-ai | 2mo | A statistical approach to bias in zero-shot learning: the lens of handwriting recognition | Generalized zero-shot learning (GZSL) has emerged as an important paradigm for visual recognition systems that must generalize to clas… | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI produced software, cannot be protected in that way. The bloat of AI companies like Anthropic and OpenAI also means if they go under, and… | AI produced software, cannot be protected in that way. The bloat of AI companies like Anthropic and OpenAI also means if they go under… | ↗Not briefed |
bluesky bluesky-ai | 2mo | More than 10% chance AI 'could kill all humans' - Anthropic researcher | More than 10% chance AI 'could kill all humans' - Anthropic researcher | ↗Not briefed |
bluesky bluesky-ai | 2mo | so many people give a pass to llm use for programming and not for visual art and audio, but seemingly forget that the thing generating the … | so many people give a pass to llm use for programming and not for visual art and audio, but seemingly forget that the thing generating… | ↗Not briefed |
rss techcrunch-ai | 2mo | Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up🔧 Suno | As it grapples with a bevy of lawsuits, Suno said its new model, Suno v6, is not trained using music it used to train previous version… | ↗Not briefed |
rss theverge-ai | 2mo | Students who use AI generally score worse at school | Students who use AI to help them study tend to perform worse at school than those who don't, according to data from a global OECD educ… | ↗Not briefed |
bluesky bluesky-ai | 2mo | "Israel está invirtiendo millones": así están manipulando ChatGPT y otras IA para que digan lo que quieren🔧 ChatGPT | "Israel está invirtiendo millones": así están manipulando ChatGPT y otras IA para que digan lo que quieren | ↗Not briefed |
hn hn-ai | 2mo | Show HN: K7 – self-hosted media server I built after years on Plex/Jellyfin | — | ↗Not briefed |
hn hn-ai | 2mo | A computer-assisted Lean 4 proof of the Berge–Fulkerson C(24) theorem | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Anthropic and OpenAI are valued at nearly a trillion dollars each and can be relied on to break all their own rules when they want to claim… | Anthropic and OpenAI are valued at nearly a trillion dollars each and can be relied on to break all their own rules when they want to… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I think one of the easiest ways to spot Anthropic's insincerity or basic cultishness is how they've actively enabled + done deals with imme… | I think one of the easiest ways to spot Anthropic's insincerity or basic cultishness is how they've actively enabled + done deals with… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Oh great BBC News - Anthropic researcher believes more than 10% chance AI 'could kill all humans' - BBC News www.bbc.co.uk/news/article... | Oh great BBC News - Anthropic researcher believes more than 10% chance AI 'could kill all humans' - BBC News www.bbc.co.uk/news/arti… | ↗Not briefed |
hn hn-ai | 2mo | Top Mathematician Announces New Institute for A.I. Safety | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | "Aloysius the Flower Boy Surprised" Scott M. Snyder (2025-2144) Painted 2058 Model, Unknown Currently Not Painted Yet. #Art #AI #Male #Homo… | "Aloysius the Flower Boy Surprised" Scott M. Snyder (2025-2144) Painted 2058 Model, Unknown Currently Not Painted Yet. #Art #AI #Male… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Physics-Informed Multi-Task Surrogate Model for the Martian Nightside Thermosphere | Modeling the Martian nightside thermosphere remains challenging due to sparse in situ sampling and strong coupling among transport, ma… | ↗Not briefed |
hn hn-ai | 2mo | Ghostscript (gstoraster): new halftone/dithering algorithms | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | LLMs make shit up but they're pretty good at catching shit other people (or other LLMs) made up. Which also means if you use fake cites, wh… | LLMs make shit up but they're pretty good at catching shit other people (or other LLMs) made up. Which also means if you use fake cite… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Tja. | Tja. | ↗Not briefed |
bluesky bluesky-ai | 2mo | With news of Anthropic not letting the UK test its latest AI model, anxiety is on the rise. Middle powers like the UK are concerned about… | With news of Anthropic not letting the UK test its latest AI model, anxiety is on the rise. Middle powers like the UK are concerned… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I won't dig much far, but we literally did the same to John Paul 2 as you did to Kirk, before LLM era. LLM or not, people would post slop, … | I won't dig much far, but we literally did the same to John Paul 2 as you did to Kirk, before LLM era. LLM or not, people would post s… | ↗Not briefed |
bluesky bluesky-ai | 2mo | ...Yo no sé si los LLM son el futuro o no, pero sé que no pueden ser dos cosas en simultaneo: La herramienta que va a remodelar el poder; y… | ...Yo no sé si los LLM son el futuro o no, pero sé que no pueden ser dos cosas en simultaneo: La herramienta que va a remodelar el pod… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Ceci dit, à force ça va finir par générer des gens persuadés que face au développement d'une machine génocidaire la seule solution est de c… | Ceci dit, à force ça va finir par générer des gens persuadés que face au développement d'une machine génocidaire la seule solution est… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Ironically, a frontier LLM can be asked "Hey, check all these references and give me a link to each cite" and then a person can check they … | Ironically, a frontier LLM can be asked "Hey, check all these references and give me a link to each cite" and then a person can check… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Olha só essa notícia. Ela saiu ontem na mídia gringa e o G1 resolveu trazer pra nossa. Preste bem atenção na chamada. - Já estava "meio" … | Olha só essa notícia. Ela saiu ontem na mídia gringa e o G1 resolveu trazer pra nossa. Preste bem atenção na chamada. - Já estava "m… | ↗Not briefed |
hn hn-ai | 2mo | The next SolarWinds will be an AI security vendor | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: A GitHub Action to stop coding agents from over-commenting | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Meet Alice and Bob. Alice works at Anthropic, Bob works at OpenAI. They're in a loving polycule. Until their partner Mal grabs both logins … | Meet Alice and Bob. Alice works at Anthropic, Bob works at OpenAI. They're in a loving polycule. Until their partner Mal grabs both lo… | ↗Not briefed |
rss techmeme | 2mo | Legal AI startup Harvey raised $550M co-led by Lightspeed and Diffusion at a $15.6B valuation, up from an $11B valuation when it raised $200M in March (Rebecca Torrence/Bloomberg) | Rebecca Torrence / Bloomberg: Legal AI startup Harvey raised $550M co-led by Lightspeed and Diffusion at a $15.6B valuation, up from a… | ↗Not briefed |
bluesky bluesky-ai | 2mo | “In addition to monitoring activists in the vicinity of Anthropic executives and keeping tabs on protests near physical Anthropic assets, t… | “In addition to monitoring activists in the vicinity of Anthropic executives and keeping tabs on protests near physical Anthropic asse… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Opair, a coding harness that eschews autonomy | Obviously everyone is writing their own coding harness these days, so I'll cut straight to the points of difference in mine:<p>*… | ↗Not briefed |