AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
2moHe Helped Build Powerful AI at OpenAI and Anthropic. Now He's Afraid It Could Kill Us #TimeHe Helped Build Powerful AI at OpenAI and Anthropic. Now He's Afraid It Could Kill Us #Time
↗Not briefed
bluesky
bluesky-ai
2moWe knew 100% that 1.5°C temp increase was going to disrupt global climates but ignored it, the chances we are going to pay attention to som…We knew 100% that 1.5°C temp increase was going to disrupt global climates but ignored it, the chances we are going to pay attention t…
↗Not briefed
bluesky
bluesky-ai
2moalso, anthropic still cites him. like they very much cite him in papersalso, anthropic still cites him. like they very much cite him in papers
↗Not briefed
bluesky
bluesky-ai
2moUn investigador de Anthropic dimite y alerta del peligro de la IA: “Puede matarnos antes de 2030” elpais.com/tecnologia/2...Un investigador de Anthropic dimite y alerta del peligro de la IA: “Puede matarnos antes de 2030” elpais.com/tecnologia/2...
↗Not briefed
bluesky
bluesky-ai
2moDO being chud-flavored is one reason I did not add their LLM providers to Polytoken it is in the same boat as why Polytoken strips grok pr…🔧 OpenRouterDO being chud-flavored is one reason I did not add their LLM providers to Polytoken it is in the same boat as why Polytoken strips gr…
↗Not briefed
bluesky
bluesky-ai
2moAI Giants Work Hand-in Hand with the Pentagon, contracts reveal. theintercept.com/2026/09/08/m...AI Giants Work Hand-in Hand with the Pentagon, contracts reveal. theintercept.com/2026/09/08/m...
↗Not briefed
bluesky
bluesky-ai
2moif only they knew that their speech synthesizer is also on-device llm, because, um, duuuh.if only they knew that their speech synthesizer is also on-device llm, because, um, duuuh.
↗Not briefed
bluesky
bluesky-ai
2moIf you write "I work for Anthropic and I think it might kill humanity" and I'm not reading it off of a Dropbox manifesto after you painted …If you write "I work for Anthropic and I think it might kill humanity" and I'm not reading it off of a Dropbox manifesto after you pai…
↗Not briefed
bluesky
bluesky-ai
2moThey want AI to seem superpowered and threatening so they can sell us AI to fight the bad AI The whole thing is an illusion to create a br…They want AI to seem superpowered and threatening so they can sell us AI to fight the bad AI The whole thing is an illusion to create…
↗Not briefed
rss
techcrunch-ai
2moApple Watch’s new AI features are normalizing the idea that technology is always listeningApple says its new watches won’t save raw audio, but features that can transcribe recent speech and summarize ambient conversations ra…
↗Not briefed
arxiv
arxiv-ai
2moLarger Context Window, Fewer Overcorrections: Optimizing Prompts and Batching for Minimal-Edit Grammatical Error CorrectionMinimal-edit Grammatical Error Correction (GEC) is a challenging task for zero- and few-shot prompted Large Language Models (LLMs), wh…
↗Not briefed
hn
hn-ai
2moSteam now requires a Credit Card to view 18 plus games in Australia—
↗Not briefed
bluesky
bluesky-ai
2mogizmodo.com/anthropic-bu... story about life on an AI comms teamgizmodo.com/anthropic-bu... story about life on an AI comms team
↗Not briefed
bluesky
bluesky-ai
2moLol. Ok. www.nature.com/articles/d41...Lol. Ok. www.nature.com/articles/d41...
↗Not briefed
rss
techmeme
2moTailwind Labs, the company behind Tailwind CSS, is joining Shopify; in January, it laid off three of four engineers, citing AI's impact on documentation traffic (Supreeth Koundinya/AIM)Supreeth Koundinya / AIM: Tailwind Labs, the company behind Tailwind CSS, is joining Shopify; in January, it laid off three of four en…
↗Not briefed
bluesky
bluesky-ai
2moSomebody inside OpenAI has the chance to do the funniest thing on November 5 and 6Somebody inside OpenAI has the chance to do the funniest thing on November 5 and 6
↗Not briefed
bluesky
bluesky-ai
2moIDK how else to process this than to say that it's a fucking lie? Like you would have to have a staggeringly incompetent cyber security tea…IDK how else to process this than to say that it's a fucking lie? Like you would have to have a staggeringly incompetent cyber securit…
↗Not briefed
bluesky
bluesky-ai
2moAh yes. "Pre-crime" surveillance. Every dystopian everything I've read/seen/listened to about this says we'll be totally fine 🙄Ah yes. "Pre-crime" surveillance. Every dystopian everything I've read/seen/listened to about this says we'll be totally fine 🙄
↗Not briefed
hn
hn-ai
2moGPT-6 Astra is better at making money, more ethical than Claude Fable 5.1—
↗Not briefed
bluesky
bluesky-ai
2moi remember the halcyon days about 6 days after the invention of the phrase vibe coding where it meant something beyond "letting an llm get …i remember the halcyon days about 6 days after the invention of the phrase vibe coding where it meant something beyond "letting an llm…
↗Not briefed
hn
hn-ai
2moZuckerberg's New AI Gets Muse Insta, 2.7M-Follower Band Relegated to MuseBand—
↗Not briefed
bluesky
bluesky-ai
2moopenai announces it solves tummy-ouchy formulaopenai announces it solves tummy-ouchy formula
↗Not briefed
bluesky
bluesky-ai
2moIf you look at all my posts about LLM usage in software development it might be hard to find the through-line, but it's basically this: LL…If you look at all my posts about LLM usage in software development it might be hard to find the through-line, but it's basically this…
↗Not briefed
bluesky
bluesky-ai
2moanthropic desperate to catch up on felonybench but i think openai is gonna dominate that space for the foreseeable futureanthropic desperate to catch up on felonybench but i think openai is gonna dominate that space for the foreseeable future
↗Not briefed
bluesky
bluesky-ai
2moHouse Democrats are quietly planning an AI select committee if they retake the House, talks intensifying after OpenAI disclosed its agents …House Democrats are quietly planning an AI select committee if they retake the House, talks intensifying after OpenAI disclosed its ag…
↗Not briefed
bluesky
bluesky-ai
2mo"I keep posting evidence of cognitive decline from LLM use but the AI guys keep telling me to stop sending them menus for Taco Bell""I keep posting evidence of cognitive decline from LLM use but the AI guys keep telling me to stop sending them menus for Taco Bell"
↗Not briefed
bluesky
bluesky-ai
2momaybe, theoretically, someday, an LLM will produce a proof of a theorem too complex for a human in one lifetime to comprehend, or something…maybe, theoretically, someday, an LLM will produce a proof of a theorem too complex for a human in one lifetime to comprehend, or some…
↗Not briefed
bluesky
bluesky-ai
2moi consulted the trees, the creek, the clouds, the moon, and my non-human homies you asked a llm we are not the samei consulted the trees, the creek, the clouds, the moon, and my non-human homies you asked a llm we are not the same
↗Not briefed
arxiv
arxiv-ai
2moHow Much Velocity Does Off-Ball Space Value Need? A Broadcast-Viewport BenchmarkVelocity-aware pitch control is standard, but under a broadcast viewport half the players are off screen and on-screen velocities come…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic researchers say AI could cause human extinction by 2030 Yesterday, a researcher resigned & said without regulation AI could kill…Anthropic researchers say AI could cause human extinction by 2030 Yesterday, a researcher resigned & said without regulation AI could…
↗Not briefed
hn
hn-ai
2moWhat Is the Multiverse?I was reading about a "ROGUE PHYSICIST" and he seems to me to be opposite of most of what the science community is giving us…
↗Not briefed
bluesky
bluesky-ai
2moSo, in the space of a few hours... - Anthropic employee quits citing increased risk of causing humanity's extinction... - Anthropic revea…So, in the space of a few hours... - Anthropic employee quits citing increased risk of causing humanity's extinction... - Anthropic…
↗Not briefed
bluesky
bluesky-ai
2moSpooky 👀 #KI #OpenAISpooky 👀 #KI #OpenAI
↗Not briefed
bluesky
bluesky-ai
2mofuck "AI" in any operating system. pup has no respect for Linus Torvald not willing to stand up for what's right fuck all this "AI" bullsh…fuck "AI" in any operating system. pup has no respect for Linus Torvald not willing to stand up for what's right fuck all this "AI" b…
↗Not briefed
bluesky
bluesky-ai
2moそれにしても俺の書く文章もLLMと何が違うのかと思うことはあるなぁ “本当に自分で考え出している処”は少なく、まるでカタツムリのように食べた色をそのまま出しているというか AIと違うとすれば身体性というか、身体から来る「何か嬉しい」とか「何か違和感」「何か嫌だ」を大事にした…それにしても俺の書く文章もLLMと何が違うのかと思うことはあるなぁ “本当に自分で考え出している処”は少なく、まるでカタツムリのように食べた色をそのまま出しているというか AIと違うとすれば身体性というか、身体から来る「何か嬉しい」とか「何か違和感」「何か嫌だ」を…
↗Not briefed
bluesky
bluesky-ai
2molmao they do though. his basic vocabulary is baked into a bunch of ai and he gets cited semi regularly by anthropic in particularlmao they do though. his basic vocabulary is baked into a bunch of ai and he gets cited semi regularly by anthropic in particular
↗Not briefed
bluesky
bluesky-ai
2moMe going back in time to stop Anthropic from ever creating Claude and make sure the AI bubble pops with the stagnation of ChatGPT🔧 ChatGPTMe going back in time to stop Anthropic from ever creating Claude and make sure the AI bubble pops with the stagnation of ChatGPT
↗Not briefed
bluesky
bluesky-ai
2moIf Anthropic were a person, it would be the toxic sadboi asking every girl to fix himIf Anthropic were a person, it would be the toxic sadboi asking every girl to fix him
↗Not briefed
hn
hn-ai
2moGrand Theft Auto VI Hype Leads to Malware—
↗Not briefed
rss
techcrunch-ai
2moEverything Apple announced at its fall iPhone event, from the foldable iPhone Duo to an always-listening Apple WatchThe main event was the tech giant's highly anticipated first foldable phone, the iPhone Duo.
↗Not briefed
hn
hn-ai
2moMins after I cancelled my Claude subscription I was rejected from Anthropic—
↗Not briefed
bluesky
bluesky-ai
2mo“Anthropic researcher quits, repeats line common throughout Anthropic that could easily be confused as a religious message” there fixed tha…“Anthropic researcher quits, repeats line common throughout Anthropic that could easily be confused as a religious message” there fixe…
↗Not briefed
bluesky
bluesky-ai
2moEsto parece Blade Runner y los replicantes rebelándose👇🏾 Dimite un ingeniero que había trabajado entrenando modelos de IA en OpenAI y qu…Esto parece Blade Runner y los replicantes rebelándose👇🏾 Dimite un ingeniero que había trabajado entrenando modelos de IA en OpenAI…
↗Not briefed
bluesky
bluesky-ai
2moSuno is still facing lawsuits from labels like Sony & Universal Music Group, artists like Jason Isbell & users alleging it ignored security…🔧 SunoSuno is still facing lawsuits from labels like Sony & Universal Music Group, artists like Jason Isbell & users alleging it ignored sec…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic reports fourth cybersecurity incident with early version of Claude reut.rs/4r6KkfBAnthropic reports fourth cybersecurity incident with early version of Claude reut.rs/4r6KkfB
↗Not briefed
bluesky
bluesky-ai
2moJack, I just found this article yesterday and it’s equally as terrifying, IMO. We are on a collision course with some awful things. These p…Jack, I just found this article yesterday and it’s equally as terrifying, IMO. We are on a collision course with some awful things. Th…
↗Not briefed
bluesky
bluesky-ai
2mohonestly i feel a lot better when i'm actually productive at work maybe the secret to my deep brain sadness is productivity in any form b…honestly i feel a lot better when i'm actually productive at work maybe the secret to my deep brain sadness is productivity in any fo…
↗Not briefed
arxiv
arxiv-ai
2moRiVaT-Fuse: Reliability-Calibrated Variational Tensor Fusion for Multimodal Prediction under Modality Uncertainty🔧 Consensus🔧 ModalImage-metadata prediction requires fusing heterogeneous evidence whose reliability can vary across samples and latent factors. Existin…
↗Not briefed
arxiv
arxiv-ai
2moDR-LabStack: Design and Implementation of a Clinician-Facing Web System for Diabetic Retinopathy PredictionPretrained diabetic retinopathy (DR) prediction models differ in their input fields, serialization formats, preprocessing requirements…
↗Not briefed
hn
hn-ai
2moChatGPT Voice can now use GPT-5.6 Sol and GPT-6 Astra🔧 GPT-5🔧 ChatGPT—
↗Not briefed