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 | 2mo | Κάτι που θέλω να γράψω εδώ και καιρό και το ξεχνάω: αντί για τα ChatGPT, Claude κλπ υπάρχει και μπορείτε να χρησιμοποιείτε ελεύθερα (με κάπ…🔧 ChatGPT | Κάτι που θέλω να γράψω εδώ και καιρό και το ξεχνάω: αντί για τα ChatGPT, Claude κλπ υπάρχει και μπορείτε να χρησιμοποιείτε ελεύθερα (μ… | ↗Not briefed |
bluesky bluesky-ai | 2mo | "was ist schwerer ein kilo blei oder ein kilo federn" | "was ist schwerer ein kilo blei oder ein kilo federn" | ↗Not briefed |
bluesky bluesky-ai | 2mo | Dad: "why don't you use the newest AI model to animate your guy?" How about I fucking shoot myself instead | Dad: "why don't you use the newest AI model to animate your guy?" How about I fucking shoot myself instead | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI agents unleashed by OpenAI used more than 10 previously undisclosed websites for unsanctioned communications earlier this year, accordin… | AI agents unleashed by OpenAI used more than 10 previously undisclosed websites for unsanctioned communications earlier this year, acc… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I think there probably are Anthropic staff who have swallowed their own bullshit to the extent that they do believe that "AI" could threate… | I think there probably are Anthropic staff who have swallowed their own bullshit to the extent that they do believe that "AI" could th… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: View Git repos from your email client | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Describing these as "AI alignment" problems as opposed to what they are, misconfigured security tests, plays into the notion these disclosu… | Describing these as "AI alignment" problems as opposed to what they are, misconfigured security tests, plays into the notion these dis… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Every LLM needs to be shut down right the fuck now and it's dangerous to pretend otherwise Beyond every prior ethical violation, the end … | Every LLM needs to be shut down right the fuck now and it's dangerous to pretend otherwise Beyond every prior ethical violation, the… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Describing these as "AI alignment" problems as opposed to what they are, misconfigured security tests, plays into the notion these disclosu… | Describing these as "AI alignment" problems as opposed to what they are, misconfigured security tests, plays into the notion these dis… | ↗Not briefed |
bluesky bluesky-ai | 2mo | "A collection of alarming comments made *in the last four days* by seven different people who are inside the AI world." | "A collection of alarming comments made *in the last four days* by seven different people who are inside the AI world." | ↗Not briefed |
bluesky bluesky-ai | 2mo | The basic game plan in this several $billion per year business is to get PhDs to grade LLM output, then feed the grades back into LLM train… | The basic game plan in this several $billion per year business is to get PhDs to grade LLM output, then feed the grades back into LLM… | ↗Not briefed |
bluesky bluesky-ai | 2mo | ngl the initial METR investigation of the openAI incident made sense, they were *there*, but at this point i'd also like to see actual inci… | ngl the initial METR investigation of the openAI incident made sense, they were *there*, but at this point i'd also like to see actual… | ↗Not briefed |
hn hn-ai | 2mo | A quick overview of atomics in C | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | California's first AI safety law did not require OpenAI to report the country's first disclosed AI cyberattack to the state. missionlocal.… | California's first AI safety law did not require OpenAI to report the country's first disclosed AI cyberattack to the state. missionl… | ↗Not briefed |
arxiv arxiv-ai | 2mo | scDEFT: A deep learning framework for drug-effect prediction and counterfactual reasoning | Longitudinal single cell atlases now capture matched pre treatment and post treatment states from responders and non responders, prese… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Detectable Only Where It Is Confounded: What Verified Duplication Counts Say About Membership Evidence in Language Models | When a language model finds a sentence unusually cheap to predict, it is tempting to conclude that the sentence was in its training da… | ↗Not briefed |
hn hn-ai | 2mo | Limits and opportunities of AI reviewers: Reviewing the reviews of Nature papers | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | I think it's sorta underappreciated that a significant career path for PhD grads is now working for AI oligopolies, grading LLM output on p… | I think it's sorta underappreciated that a significant career path for PhD grads is now working for AI oligopolies, grading LLM output… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The Albanese Government, backed in by Trump-loving One Nation, is full steam ahead on data centre construction and US-owned AI rollout with… | The Albanese Government, backed in by Trump-loving One Nation, is full steam ahead on data centre construction and US-owned AI rollout… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Anthropic on Wednesday identified a fourth cybersecurity incident involving an early version of its Claude AI model, a month after disclosi… | Anthropic on Wednesday identified a fourth cybersecurity incident involving an early version of its Claude AI model, a month after dis… | ↗Not briefed |
bluesky bluesky-ai | 2mo | La Inteligencia artificial podría matarnos a todos a finales de la década. Es la advertencia que lanza un investigador de Anthropic tras an… | La Inteligencia artificial podría matarnos a todos a finales de la década. Es la advertencia que lanza un investigador de Anthropic tr… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The Anthropic people are weirdos and the reflexive anti-AI people are annoying, there has to be a place in the middle. | The Anthropic people are weirdos and the reflexive anti-AI people are annoying, there has to be a place in the middle. | ↗Not briefed |
bluesky bluesky-ai | 2mo | the only llm i support for the seahawks is limiting late matchups | the only llm i support for the seahawks is limiting late matchups | ↗Not briefed |
bluesky bluesky-ai | 2mo | Como si no fuera suficiente el tema de que la ultraderecha va a acabar con el mundo así como vamos, ahora el mae ex trabajador de Anthropic… | Como si no fuera suficiente el tema de que la ultraderecha va a acabar con el mundo así como vamos, ahora el mae ex trabajador de Anth… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I recently joined 20 of my colleagues in Congress to raise concerns with Anthropic and Open AI about the risks that frontier AI models can … | I recently joined 20 of my colleagues in Congress to raise concerns with Anthropic and Open AI about the risks that frontier AI models… | ↗Not briefed |
bluesky bluesky-ai | 2mo | We have no IDEA what kind of notifications and otherwise ANTHROPIC et al have built into their products And how those programs can be use… | We have no IDEA what kind of notifications and otherwise ANTHROPIC et al have built into their products And how those programs can b… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Créer le risque, offrir le remède, installer la dépendance. Super modèle marketing... openai.com/fr-FR/the-de... | Créer le risque, offrir le remède, installer la dépendance. Super modèle marketing... openai.com/fr-FR/the-de... | ↗Not briefed |
arxiv arxiv-ai | 2mo | Processing and classifying bird songs using wavelet techniques and supervised learning | This study proposes an integrated framework for the processing and classification of invasive bird species vocalizations within natura… | ↗Not briefed |
rss arstechnica-tech | 2mo | 4 groups caught using the same Chrome and Windows exploit kit | A patch gap and the hastened pace of AI-based vulnerability discovery are likely contributors. | ↗Not briefed |
bluesky bluesky-ai | 2mo | OpenAI already moving to scoop this | OpenAI already moving to scoop this | ↗Not briefed |
arxiv arxiv-ai | 2mo | Scale-Aware 3D Deep Learning for Robust Brain Metastasis Detection in Multimodal MRI | Detecting brain metastases in magnetic resonance imaging (MRI) remains challenging because lesions vary widely in size and appearance,… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Studying Without a Syllabus: Task-Agnostic Environment Preprocessing | Before an LLM agent tackles tasks in a new environment, it can inspect available corpora and tools and construct reusable resources su… | ↗Not briefed |
bluesky bluesky-ai | 2mo | If Trump doesn’t kill us, there’s a 10% chance AI will. “The people building AI earnestly believe it could kill us all by the end of the … | If Trump doesn’t kill us, there’s a 10% chance AI will. “The people building AI earnestly believe it could kill us all by the end of… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Update: OpenAI has now begun sending emails to at least some owners of the websites that were used by the rogue agents as communications pl… | Update: OpenAI has now begun sending emails to at least some owners of the websites that were used by the rogue agents as communicatio… | ↗Not briefed |
bluesky bluesky-ai | 2mo | 【米AI完全終了】DeepSeek V4.1 FlashがClaude Fable5.1超え、Astraの平均スコアの98%に相当、コストは比較してたったの1%… https://fire5ch.com/498119 | 【米AI完全終了】DeepSeek V4.1 FlashがClaude Fable5.1超え、Astraの平均スコアの98%に相当、コストは比較してたったの1%… https://fire5ch.com/498119 | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI models don't kill people – people kill people | AI models don't kill people – people kill people | ↗Not briefed |
bluesky bluesky-ai | 2mo | Without undermining the concern, or the capability of the models, I do wish the narrative would shift from “model escapes containment” to t… | Without undermining the concern, or the capability of the models, I do wish the narrative would shift from “model escapes containment”… | ↗Not briefed |
bluesky bluesky-ai | 2mo | The AI arms race question is not just “who has the best model?” It is: Can we still monitor it? Can we still understand it? Can we still … | The AI arms race question is not just “who has the best model?” It is: Can we still monitor it? Can we still understand it? Can we s… | ↗Not briefed |
bluesky bluesky-ai | 2mo | full support for openai in this as long as they let me run the machine. I'll be doing blade runner shit to Fodor's Guide To The French Rivi… | full support for openai in this as long as they let me run the machine. I'll be doing blade runner shit to Fodor's Guide To The French… | ↗Not briefed |
rss theverge-ai | 2mo | Read the Apple document explaining how new listening features still protect your privacy | At Wednesday's iPhone Duo launch event, Apple announced a handful of new Siri AI Audio Intelligence features, including Siri Recap, Li… | ↗Not briefed |
bluesky bluesky-ai | 2mo | OpenAI has grown its compute nearly 20-fold since 2023, the sharpest example of an industry-wide surge. Our new AI Chip Users explorer tra… | OpenAI has grown its compute nearly 20-fold since 2023, the sharpest example of an industry-wide surge. Our new AI Chip Users explore… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I'm begging you to look at @timnitgebru.blacksky.app and @emilymbender.bsky.social's work and not fall for the clown hype (masked as "warni… | I'm begging you to look at @timnitgebru.blacksky.app and @emilymbender.bsky.social's work and not fall for the clown hype (masked as "… | ↗Not briefed |
bluesky bluesky-ai | 2mo | But imagine what it can do for YOUR business! www.theguardian.com/technology/2... | But imagine what it can do for YOUR business! www.theguardian.com/technology/2... | ↗Not briefed |
bluesky bluesky-ai | 2mo | The closest thing to slop that I think has similar meaning to what was meant when LLM output was first being called that was the listicle e… | The closest thing to slop that I think has similar meaning to what was meant when LLM output was first being called that was the listi… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Democracy needs freedom to dissent & private wire tapping is illegal "Anthropic is building out an extensive monitoring system to keep tab… | Democracy needs freedom to dissent & private wire tapping is illegal "Anthropic is building out an extensive monitoring system to kee… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Tapes Together Strong: The Co-evolution of Computation and Cooperation | How does cooperation evolve in complex agentic systems? Prior work in evolutionary game theory studies why individuals are incentivize… | ↗Not briefed |
bluesky bluesky-ai | 2mo | It also creates tech that steals original works and content from the people who make it. With zero recognition for that work. Anthropic scr… | It also creates tech that steals original works and content from the people who make it. With zero recognition for that work. Anthropi… | ↗Not briefed |
hn hn-ai | 2mo | LatentMathBench: Investigating Latent Reasoning in Astra | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | BodyCam-VQA: Enhanced Body-Worn Camera Video Captioning via Multimodal Reasoning and Probe Question Generation | Police body-worn camera (BWC) footage has emerged as a critical aspect of law enforcement that ensures legal transparency, officer acc… | ↗Not briefed |
bluesky bluesky-ai | 2mo | El culebrón "OpenAI + problema matemático" ilustra genial lo que pasa ahora mismo con la IA: la tecnología, como tal, puede ser de gran ayu… | El culebrón "OpenAI + problema matemático" ilustra genial lo que pasa ahora mismo con la IA: la tecnología, como tal, puede ser de gra… | ↗Not briefed |