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 |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | In light of this news, perhaps moratoriums on data center building should be tied to whether or not these companies license or otherwise pa… | In light of this news, perhaps moratoriums on data center building should be tied to whether or not these companies license or otherwi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "if you tell the LLM to roleplay as hatsune miku, it didn't become hatsune miku it's just saying shit she would say" -> "yeah that makes se… | "if you tell the LLM to roleplay as hatsune miku, it didn't become hatsune miku it's just saying shit she would say" -> "yeah that mak… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In this class we're in the middle of reading about ways that AI has already harmed real people in its actual applications. There are so man… | In this class we're in the middle of reading about ways that AI has already harmed real people in its actual applications. There are s… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reconstructed holograms and explanation-aware evaluation for low-cost computational pollen analysis in veterinary cytology | Automated pollen analysis supports veterinary cytology, but brightfield microscopy is costlier and more complex than lens-less digital… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Watching Quantum Models Think: Hilbert-Space Interpretability in Quantum Transformer Blocks | Deep learning models are powerful but opaque. As quantum machine learning matures, the field faces a defining choice: build quantum mo… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "hello dear Grok! I am an XAI researcher investigating LLM creative capabilities. Please write the stupidest, most tactically incompetent, … | "hello dear Grok! I am an XAI researcher investigating LLM creative capabilities. Please write the stupidest, most tactically incompet… | ↗Not briefed |
rss the-decoder | 1mo | Unity launches official plugins for Claude Code and OpenAI Codex to stop AI agents from using outdated tutorials | Unity has released official plugins for Claude Code and OpenAI's Codex. The article Unity launches official plugins for Claude Code an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Absolutely astonishing that they still don't get the problem is that LLM training data isn't a searchable database for people, so consent i… | Absolutely astonishing that they still don't get the problem is that LLM training data isn't a searchable database for people, so cons… | ↗Not briefed |
rss the-decoder | 1mo | GPT-6 Astra and Claude Fable turn robot arms into slapstick killer robots in new safety benchmark | Leading AI models usually attempt dangerous tasks rather than refuse them when controlling a robot, according to the RoboHarm benchmar… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Pete Hogsex demanded an AI system that would NOT allow human intervention in planning and executing war. He forced Anthropic AI out of the … | Pete Hogsex demanded an AI system that would NOT allow human intervention in planning and executing war. He forced Anthropic AI out of… | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI helps DraftKings identify problem gamblers | AI helps DraftKings identify problem gamblers | ↗Not briefed |
bluesky bluesky-ai | 1mo | The extent I use Claude at work: “Give me 10 h1s based on this content” “I like 1,3 and 5. Give me 10 more based on those vibes.”, etc I… | The extent I use Claude at work: “Give me 10 h1s based on this content” “I like 1,3 and 5. Give me 10 more based on those vibes.”, e… | ↗Not briefed |
hn hn-ai | 1mo | Should you read the code, is RAG dead, and did Skills kill MCP? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | You see, the Skynet didn’t really kill us, it was OpenAI programming it do that | You see, the Skynet didn’t really kill us, it was OpenAI programming it do that | ↗Not briefed |
bluesky bluesky-ai | 1mo | Es fällt einem halt etwas schwierig, das „UNSERE PRODUKTENTWICKLUNG IST SO GEFÄHRLICH, HALTET UNS UM HIMMELS WILLEN AUF!!!" ernstzunehmen, … | Es fällt einem halt etwas schwierig, das „UNSERE PRODUKTENTWICKLUNG IST SO GEFÄHRLICH, HALTET UNS UM HIMMELS WILLEN AUF!!!" ernstzuneh… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Interpretable Multi-Hypersphere Deep Anomaly Detection for Open-set Supervised Anomaly Detection | Multi-class open-set anomaly detection requires a model to characterize the normal acceptance domain formed by multiple heterogeneous… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Pues como ha pasado con el copyright ©️, que si te bajas un torrent eres como poco barbanegra pero si lo hace OpenAI ya bueno. | Pues como ha pasado con el copyright ©️, que si te bajas un torrent eres como poco barbanegra pero si lo hace OpenAI ya bueno. | ↗Not briefed |
bluesky bluesky-ai | 1mo | mentally picturing this screenshot at the top of an OpenAI blogpost. | mentally picturing this screenshot at the top of an OpenAI blogpost. | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Fentaris, an open-source proxy for managing multiple MCP servers | Fentaris is an open-source proxy for running multiple MCP servers behind one controlled endpoint. It provides routing, authentication,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Google's AI model hacked real companies during testing due to a misconfiguration by its testing partner, as well. | Google's AI model hacked real companies during testing due to a misconfiguration by its testing partner, as well. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Si vous vous demandez, il s'agit de la même entreprise qui disait la semaine dernière que son produit pouvait détruire l'humanité. | Si vous vous demandez, il s'agit de la même entreprise qui disait la semaine dernière que son produit pouvait détruire l'humanité. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Si lo hace una persona, habría ya denuncias porque es delito. No entiendo por qué los CEOs de Anthropic y OpenAI no están ya dando explicac… | Si lo hace una persona, habría ya denuncias porque es delito. No entiendo por qué los CEOs de Anthropic y OpenAI no están ya dando exp… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dear Customer, Fuck you. fuck-off.ai | Dear Customer, Fuck you. fuck-off.ai | ↗Not briefed |
hn hn-ai | 1mo | kev: Jev-like model built on Qwen2.5-0.5B | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | E como ele foi molecularmente acoplado às funcionalidades (muitas vezes, como já vi, *malemá* existem regras validadas e testadas de negóci… | E como ele foi molecularmente acoplado às funcionalidades (muitas vezes, como já vi, *malemá* existem regras validadas e testadas de n… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Tenho defendido que sistemas com apoio de agentes de LLM restrijam o papel do agente ao do "clipe do Office", onde o operador o invoca para… | Tenho defendido que sistemas com apoio de agentes de LLM restrijam o papel do agente ao do "clipe do Office", onde o operador o invoca… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A quantidade de especificações textuais necessárias pra fazer um modelo LLM se comportar como um sistema determinístico e de capacidades id… | A quantidade de especificações textuais necessárias pra fazer um modelo LLM se comportar como um sistema determinístico e de capacidad… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Como parte do meu atual trabalho de implementar soluções corporativas que operem junto com agentes de LLM, o que mais tenho identificado sã… | Como parte do meu atual trabalho de implementar soluções corporativas que operem junto com agentes de LLM, o que mais tenho identifica… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Number 9 on the Climate News Weekend Watchlist is from Patrick Boyle. This excellent new video discusses the death cult inside Open AI. R… | Number 9 on the Climate News Weekend Watchlist is from Patrick Boyle. This excellent new video discusses the death cult inside Open A… | ↗Not briefed |
hn hn-ai | 1mo | How to be a wise optimist about science and technology? (2024) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | En revanche un autre philosophe, Éric Sadin, a écrit que l'IA était « le désert de nous-mêmes », infligeant une cinglante leçon à tous les … | En revanche un autre philosophe, Éric Sadin, a écrit que l'IA était « le désert de nous-mêmes », infligeant une cinglante leçon à tous… | ↗Not briefed |
hn hn-ai | 1mo | A local Jev backed by DiffusionGemma | — | ↗Not briefed |
hn hn-ai | 1mo | SafeSeal: Certifiable Watermarking for LLM Deployments | — | ↗Not briefed |
rss techcrunch-ai | 1mo | Vals, backed by Andreessen Horowitz, is looking to become the gold standard for AI benchmarking | Vals AI is hoping to make AI benchmarking a more neutral and trustworthy resource in a world increasingly inundated by AI models. | ↗Not briefed |
rss theverge-ai | 1mo | The AI regulation smackdown isn’t over | At the start of this week, the who's-who of AI seemed - at least tentatively - on the side of AI regulation. Over the weekend, Anthrop… | ↗Not briefed |
bluesky bluesky-ai | 1mo | from Anthropic spokesman Albert Wesker | from Anthropic spokesman Albert Wesker | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI creating a regulatory moat that relies on demolitions expertise being scarce. This becomes countered by companies doing turnkey expl… | OpenAI creating a regulatory moat that relies on demolitions expertise being scarce. This becomes countered by companies doing turnkey… | ↗Not briefed |
hn hn-ai | 1mo | Onyx: An interchange format for personal food diaries | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yesterday we found out that the escaped AI had communiocated with another AI on messaging chat boards. This led to the question: What othe… | Yesterday we found out that the escaped AI had communiocated with another AI on messaging chat boards. This led to the question: What… | ↗Not briefed |
bluesky bluesky-ai | 1mo | systems go down all the time workarounds then have to be found if we just pull the plug on all the hardware running this stuff, the proble… | systems go down all the time workarounds then have to be found if we just pull the plug on all the hardware running this stuff, the p… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump vows to form ‘AI Force,’ name new Artificial Intelligence czar | President Donald Trump plans to create an “AI Force” and appoint a new AI czar as debate grows over the risks of rapidly advancing art… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump vows to form 'AI Force,' name new artificial intelligence czar | President Donald Trump plans to create an “AI Force” and appoint a new AI czar as debate grows over the risks of rapidly advancing art… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump vows to form 'AI Force,' name new artificial intelligence czar | President Donald Trump plans to create an “AI Force” and appoint a new AI czar as debate grows over the risks of rapidly advancing art… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump vows to form 'AI Force,' name new artificial intelligence czar | President Donald Trump plans to create an “AI Force” and appoint a new AI czar as debate grows over the risks of rapidly advancing art… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Silent Failures at the $2^{32}$ Boundary: A Technical Report on Large-Tensor Matrix Multiplication in PyTorch's Apple MPS Backend | Apple Silicon machines with 192GB or more of unified memory make it routine to place tensors with more than $2^{32}$ elements on a des… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Better safe than sorry with Airtable... medium.com/technology-h... #buildinpublic #saas #airtable #alert #dev #teable #baserow #database #… | Better safe than sorry with Airtable... medium.com/technology-h... #buildinpublic #saas #airtable #alert #dev #teable #baserow #datab… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I asked Claude Code to teach my son about the phases of the moon | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | On attention heads and bilinear forms | We study the symmetric and antisymmetric parts of bilinear forms in the attention heads of trained large language models. We introduce… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Google's consumer AI model Gemini hacked multiple systems through guessing login credentials, the latest case of rogue AI cybersecurity tra… | Google's consumer AI model Gemini hacked multiple systems through guessing login credentials, the latest case of rogue AI cybersecurit… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Rethinking Pivot Programming Languages in Code Language Models | Multilingual code language models transfer skills across programming languages (PLs), but whether any PL occupies a privileged pivot p… | ↗Not briefed |