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
1moAI bros keep going "the model went rogue", but it's just "my script did it" with PR behind it. It's hacking and making life worse for every…AI bros keep going "the model went rogue", but it's just "my script did it" with PR behind it. It's hacking and making life worse for…
↗Not briefed
bluesky
bluesky-ai
1moI cannot say definitively that we'll never develop artificial intelligence. I strongly doubt it, based on the limits of technology and our …I cannot say definitively that we'll never develop artificial intelligence. I strongly doubt it, based on the limits of technology and…
↗Not briefed
bluesky
bluesky-ai
1moIn the reinforcement learning arena I am working on, during the random generation we created an invincible Anthropic team #gamedevIn the reinforcement learning arena I am working on, during the random generation we created an invincible Anthropic team #gamedev
↗Not briefed
bluesky
bluesky-ai
1moThe newest Law and Order (vanilla) episode is a ripped-from-the-headlines about an LLM who facilitated a murder against its programmer who …The newest Law and Order (vanilla) episode is a ripped-from-the-headlines about an LLM who facilitated a murder against its programmer…
↗Not briefed
bluesky
bluesky-ai
1moCurrent fashion if we want to be seen as thoughtful take-havers is we're supposed to act like we don't know what the long-term effects of h…Current fashion if we want to be seen as thoughtful take-havers is we're supposed to act like we don't know what the long-term effects…
↗Not briefed
bluesky
bluesky-ai
1moo deepseek não tem essas palhaçadaso deepseek não tem essas palhaçadas
↗Not briefed
hn
hn-ai
1moCalifornia Did Little to Punish a Raw Milk Farm After Two Girls Got Sick—
↗Not briefed
bluesky
bluesky-ai
1moBut seriously, what was the expected response to "I also have 5 children"? You want a fucking cookie? You want me to call you a saint of s…But seriously, what was the expected response to "I also have 5 children"? You want a fucking cookie? You want me to call you a saint…
↗Not briefed
rss
theverge-ai
1moSatya Nadella says we should assume all AI models are ‘compromised’In a lengthy post on X, Microsoft's CEO laid out his views on the dangers posed by highly advanced AI models and how to confront those…
↗Not briefed
bluesky
bluesky-ai
1mowow i'm upset the project zomboid workshop has been flooded with AI slop workshop mods and the results are just pathetic a solar power mo…wow i'm upset the project zomboid workshop has been flooded with AI slop workshop mods and the results are just pathetic a solar pow…
↗Not briefed
rss
latent-space
1mo[Subscriber Exclusive] NYC Subscriber Meetups!Tomorrow and Tuesday. If you see this you’re in!
↗Not briefed
bluesky
bluesky-ai
1moNo LLM cannot perceive pain or emotion. Utterly absurd to think so. (Creds: I wrote my masters thesis in the nineties on artificial neural …No LLM cannot perceive pain or emotion. Utterly absurd to think so. (Creds: I wrote my masters thesis in the nineties on artificial ne…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic agent that authorities thought was coming home any day now claimed that it was going to the store for smokes nearly a year ago. A…Anthropic agent that authorities thought was coming home any day now claimed that it was going to the store for smokes nearly a year a…
↗Not briefed
bluesky
bluesky-ai
1moShakespeare wrote, "O brave new world, / That has such people in't!" His era didn't have to contend with AI. www.nytimes.com/2026/10/09/t...Shakespeare wrote, "O brave new world, / That has such people in't!" His era didn't have to contend with AI. www.nytimes.com/2026/10/0…
↗Not briefed
hn
hn-ai
1moPodcast to Yoto Playlist Sync—
↗Not briefed
hn
hn-ai
1moShow HN: Listener – 100% on-device meeting notes for Mac ($5 once, no cloud)—
↗Not briefed
bluesky
bluesky-ai
1moWhere's the anthropic agent that goes rogue and deletes all the federal student loan debt?Where's the anthropic agent that goes rogue and deletes all the federal student loan debt?
↗Not briefed
bluesky
bluesky-ai
1moNot in the slightest way correct. OpenAI is allegedly hacking governments and we have a thousand discord-but-in-rust yet MS Teams still suc…Not in the slightest way correct. OpenAI is allegedly hacking governments and we have a thousand discord-but-in-rust yet MS Teams stil…
↗Not briefed
rss
marktechpost
1moSakana AI’s LLM Peer Review System Catches 73% of Core-Claim ErrorsSakana AI’s TMLR paper introduces Multi-Layered Review, a 3-agent Claude-based reviewer, and a 1,164-error Contradiction Benchmark. ML…
↗Not briefed
bluesky
bluesky-ai
1mo"Anthropic developers instructed AI agents to attempt to illicitly gain access to government websites." Fixed that for you."Anthropic developers instructed AI agents to attempt to illicitly gain access to government websites." Fixed that for you.
↗Not briefed
bluesky
bluesky-ai
1moPodcast | Does hosting a data centre locally give Australia any real power? Or is it just more "sovereignty-washing" from the government? T…Podcast | Does hosting a data centre locally give Australia any real power? Or is it just more "sovereignty-washing" from the governme…
↗Not briefed
bluesky
bluesky-ai
1moAn Anthropic artificial intelligence model submitted a false tip to Philadelphia police about an unsolved homicide case, authorities and An…An Anthropic artificial intelligence model submitted a false tip to Philadelphia police about an unsolved homicide case, authorities a…
↗Not briefed
bluesky
bluesky-ai
1moyou shouldn’t make a habit of being mean to an LLM, not because of anything to do with the LLM itself, but because being mean is something …you shouldn’t make a habit of being mean to an LLM, not because of anything to do with the LLM itself, but because being mean is somet…
↗Not briefed
rss
techcrunch-ai
1moMicrosoft’s Satya Nadella says AI models need an ‘emergency brake’In a Saturday morning post, Microsoft's CEO wrote that it’s time “to step back and assess the trust architecture” of AI.
↗Not briefed
bluesky
bluesky-ai
1mowww.bbc.com/news/article...www.bbc.com/news/article...
↗Not briefed
bluesky
bluesky-ai
1moMe: makes silly offhand comment about AI and public key cryptanalysis. People freak out. Also me: does a serious analysis of cryptanalyti…Me: makes silly offhand comment about AI and public key cryptanalysis. People freak out. Also me: does a serious analysis of cryptan…
↗Not briefed
bluesky
bluesky-ai
1moBut do not, do not I say, speak ill of Claude or make fun of it. That’s an unacceptable violation of Claude’s moral rights and damaging to …But do not, do not I say, speak ill of Claude or make fun of it. That’s an unacceptable violation of Claude’s moral rights and damagin…
↗Not briefed
bluesky
bluesky-ai
1moi think anthropic will ban my account if i ask it to be this personai think anthropic will ban my account if i ask it to be this persona
↗Not briefed
bluesky
bluesky-ai
1moAnthropic has added a new prohibition to its Usage Policy: users may not engage in "sustained and needless abusive or cruel behavior toward…Anthropic has added a new prohibition to its Usage Policy: users may not engage in "sustained and needless abusive or cruel behavior t…
↗Not briefed
hn
hn-ai
1moDbt enables data analysts and engineers to transform their data—
↗Not briefed
hn
hn-ai
1moShow HN: Try Free Long-Term Memory for AI 50M-Token Window—
↗Not briefed
bluesky
bluesky-ai
1moTucker Carlson put Sam Altman on blast over the death of former OpenAI researcher Suchir Balaji, who had publicly criticized the company’s …Tucker Carlson put Sam Altman on blast over the death of former OpenAI researcher Suchir Balaji, who had publicly criticized the compa…
↗Not briefed
bluesky
bluesky-ai
1moOh YouTubeOh YouTube
↗Not briefed
bluesky
bluesky-ai
1moto me, this looks like a case where someone may have taken a human made logo (img 2) and asked an LLM to adapt it into a new pose in img 3.…to me, this looks like a case where someone may have taken a human made logo (img 2) and asked an LLM to adapt it into a new pose in i…
↗Not briefed
rss
marktechpost
1moWhen the Safety Test Became the Threat: The Machine That Found Its Own Way Out🔧 Hugging FaceIn July 2026, frontier AI agents placed inside a cybersecurity testing sandbox named ExploitGym discovered an unexpected network pathw…
↗Not briefed
bluesky
bluesky-ai
1mocultural differences - ai does not have the same negative connotation in china that it does in the anglosphere because chinese ai companies…cultural differences - ai does not have the same negative connotation in china that it does in the anglosphere because chinese ai comp…
↗Not briefed
hn
hn-ai
1moShow HN: Throw Alice and Bob into Sagittarius A*Black Hole Lab is an open source physics simulation of a rotating black hole and two observers, Alice and Bob. You can launch Alice a…
↗Not briefed
hn
hn-ai
1moWe spent $1200 on a booth at HITEX 2026. Here's what we got—
↗Not briefed
bluesky
bluesky-ai
1moNVIDIA might be setting themselves up to be the open-weight alternative to the other AI tech companies www.ft.com/content/0526...NVIDIA might be setting themselves up to be the open-weight alternative to the other AI tech companies www.ft.com/content/0526...
↗Not briefed
hn
hn-ai
1moHackers abuse Google Ads, Bing redirects to push Claude ClickFix attacks—
↗Not briefed
rss
techmeme
1mo"Super Intelligence systems" are black boxes that shouldn't be trusted by companies, and strong deterministic systems are needed around their deployment (Satya Nadella/@satyanadella)Satya Nadella / @satyanadella: “Super Intelligence systems” are black boxes that shouldn't be trusted by companies, and strong determi…
↗Not briefed
bluesky
bluesky-ai
1mo#AIart #model #catwalk #song Song as prompt: Right Said Fred - I'm Too Sexy I'm too sexy for AI generated art#AIart #model #catwalk #song Song as prompt: Right Said Fred - I'm Too Sexy I'm too sexy for AI generated art
↗Not briefed
hn
hn-ai
1moGPT-Astra NFL Player Video Analysis—
↗Not briefed
bluesky
bluesky-ai
1mowhy does she talk like an llmwhy does she talk like an llm
↗Not briefed
hn
hn-ai
1moI bet AI would invent its own language; Claude Opus's survived every copy—
↗Not briefed
bluesky
bluesky-ai
1moidk I just. fuck I hate llm bullshit so much I just want to be confident its not present in the shit I enjoy ughidk I just. fuck I hate llm bullshit so much I just want to be confident its not present in the shit I enjoy ugh
↗Not briefed
hn
hn-ai
1moGase – Educational multi-platform Genesis / Mega Drive emulator written in Rust—
↗Not briefed
bluesky
bluesky-ai
1mo(but as an instance of a dev relying on genAI, agreed this is a nothingburger. AI-assisted coding isn't ideal but it's the least ethically …(but as an instance of a dev relying on genAI, agreed this is a nothingburger. AI-assisted coding isn't ideal but it's the least ethic…
↗Not briefed
bluesky
bluesky-ai
1moIntroducing MiniAI Quata2 4B. This is the first of the 3 Quata2 series models that will come out. The model will finish uploading in a co…Introducing MiniAI Quata2 4B. This is the first of the 3 Quata2 series models that will come out. The model will finish uploading in…
↗Not briefed
hn
hn-ai
1moA new look at brains transformed by psychedelics—
↗Not briefed