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
hn
hn-ai
11dSydney researchers uncover a hidden limit to AI predictability—
↗Not briefed
bluesky
bluesky-ai
11dSomething that strikes me about anthropic kneecapping the slop "recompilation" games is that it's a phenomenal example of how this tech is …Something that strikes me about anthropic kneecapping the slop "recompilation" games is that it's a phenomenal example of how this tec…
↗Not briefed
bluesky
bluesky-ai
11dYeah, they definitely should have disclosed. I'm personally not bothered by the LLM usage itself, but am bothered by the lack of disclosure…Yeah, they definitely should have disclosed. I'm personally not bothered by the LLM usage itself, but am bothered by the lack of discl…
↗Not briefed
hn
hn-ai
11dUkraine's Fiber-Optic Drones Compete with Data Centers for Optical Fiber Supply—
↗Not briefed
hn
hn-ai
11dShow HN: pvewarden – let an AI agent patch Proxmox VMs without giving it a shell—
↗Not briefed
hn
hn-ai
11dClaude Fable 5.2, one day later [video]—
↗Not briefed
bluesky
bluesky-ai
11dYes, but at the same time the lack of transparency until it mattered is what most have an issue with. I think building a nuanced stance for…Yes, but at the same time the lack of transparency until it mattered is what most have an issue with. I think building a nuanced stanc…
↗Not briefed
hn
hn-ai
11dShow HN: Kafuck – Apache Kafka Rewritten in Brainfuck—
↗Not briefed
arxiv
arxiv-ai
11dWhat a CAGE 4 Score Measures: An Audit of a Learned Cyber-Defense Benchmark, Its Reference Defenders, and the Tables Built on ThemCAGE Challenge 4 is the benchmark on which learned multi-agent cyber defenders are compared. We reproduce its two reference defenders,…
↗Not briefed
bluesky
bluesky-ai
11dAnyway, before you make an opinion, I do think it's important to read their own disclosure regarding their LLM usage here: bigwalk.game/faq…Anyway, before you make an opinion, I do think it's important to read their own disclosure regarding their LLM usage here: bigwalk.gam…
↗Not briefed
bluesky
bluesky-ai
11dWhile I remain an advocate against the normalisation or frivolous use of AI, I also recognize the nuance between "we were stuck at the end …While I remain an advocate against the normalisation or frivolous use of AI, I also recognize the nuance between "we were stuck at the…
↗Not briefed
bluesky
bluesky-ai
11dIt's always "harmless" if its not immediately visible, but optimization, problem solving, future-proofing? Those are things you get with pr…It's always "harmless" if its not immediately visible, but optimization, problem solving, future-proofing? Those are things you get wi…
↗Not briefed
bluesky
bluesky-ai
11dY'all should read the article, because they used an LLM to fix a single bug they couldn't figure out a fix for even after paying a bunch of…Y'all should read the article, because they used an LLM to fix a single bug they couldn't figure out a fix for even after paying a bun…
↗Not briefed
bluesky
bluesky-ai
11dI’m still gonna finish my project bc I’m having fun and they’re gifts for my friends and it’s more about celebrating that than the game its…I’m still gonna finish my project bc I’m having fun and they’re gifts for my friends and it’s more about celebrating that than the gam…
↗Not briefed
arxiv
arxiv-ai
11dGeneralizing Causal Decision-Making: Markov Equivalence, Transportability and CounterfactualsStructural causal bandits use causal knowledge to guide exploration when reward mechanisms are unknown. This paper extends the approac…
↗Not briefed
rss
theverge-ai
11dHands-on with Ring’s new AI drone camera for your homeRing’s new Drone Cam is a flying indoor security camera. Ring is finally shipping a flying indoor drone camera. For real this time.…
↗Not briefed
hn
hn-ai
11dCan a Dutch language model support journalists while preserving publicitations—
↗Not briefed
bluesky
bluesky-ai
11dAustralia plans to create an artificial intelligence safety regulator in response to a breach of its health department website by OpenAI la…Australia plans to create an artificial intelligence safety regulator in response to a breach of its health department website by Open…
↗Not briefed
bluesky
bluesky-ai
11dSerá que ele trabalha pra Anthropic, @msoares.bsky.social ? g1.globo.com/rj/rio-de-ja...Será que ele trabalha pra Anthropic, @msoares.bsky.social ? g1.globo.com/rj/rio-de-ja...
↗Not briefed
bluesky
bluesky-ai
11dalso I read through, they specifically say they used an LLM. that's the line.also I read through, they specifically say they used an LLM. that's the line.
↗Not briefed
bluesky
bluesky-ai
11dCould deal with the new job falling through, but having an AI-generated leaving card sent round to sign for someone else at work that was m…Could deal with the new job falling through, but having an AI-generated leaving card sent round to sign for someone else at work that…
↗Not briefed
bluesky
bluesky-ai
11dkurz vor dem crash auch mal eine gute AI Nachricht: #opensource und #offenstandards stehen auch wegen AI goldenen Zeiten bevor. Und zwar n…kurz vor dem crash auch mal eine gute AI Nachricht: #opensource und #offenstandards stehen auch wegen AI goldenen Zeiten bevor. Und z…
↗Not briefed
bluesky
bluesky-ai
11di think there's plenty of interesting ML tools, but then you're talking about projects like UVR that are fully open source and that don't e…i think there's plenty of interesting ML tools, but then you're talking about projects like UVR that are fully open source and that do…
↗Not briefed
arxiv
arxiv-ai
11dWhere Do Vision-Language Models Get Left and Right? Position-Code Counterfactuals in Qwen2.5-VL, Qwen3-VL and InternVL3A vision-language model may take left-right direction from the orientation of image content within patches, its vision transformer's (…
↗Not briefed
arxiv
arxiv-ai
11dHandsOnAid: Building and Evaluating an Offline First-Aid Assistant for Emergency Response Scenarios🔧 GPT-5Emergencies can occur where there is no phone signal and no ambulance nearby. A small language model on a phone can be helpful in such…
↗Not briefed
hn
hn-ai
11dShow HN: View Markdown like you would on GitHub/Obsidian, right in your terminalHi HN! I (for better or worse) live in the terminal and look at Markdown files a lot for work.<p>To read Markdown, it seems like we ha…
↗Not briefed
bluesky
bluesky-ai
11dNasdaq futures slip with Nvidia woes, Anthropic IPO in focus reut.rs/4iOXZUUNasdaq futures slip with Nvidia woes, Anthropic IPO in focus reut.rs/4iOXZUU
↗Not briefed
bluesky
bluesky-ai
11dBig Walk is a game made with artistry and consideration. It's aberrant that they'd resort to LLM at the 11th hour to solve an issue that th…Big Walk is a game made with artistry and consideration. It's aberrant that they'd resort to LLM at the 11th hour to solve an issue th…
↗Not briefed
bluesky
bluesky-ai
11dliquid bluesky that you have hundreds of quote bleets screaming at the developers of a beloved indie game and you look it up and they used …liquid bluesky that you have hundreds of quote bleets screaming at the developers of a beloved indie game and you look it up and they…
↗Not briefed
bluesky
bluesky-ai
11d..εναρκτήριας με λίγο διαφορετικό τρόπο. Κακή αυτόματη μετάφραση η φτιαγμένο σε LLM δεν έχει σημασία, έχει χαθεί ο άνθρωπος. Εκτός από το κ…..εναρκτήριας με λίγο διαφορετικό τρόπο. Κακή αυτόματη μετάφραση η φτιαγμένο σε LLM δεν έχει σημασία, έχει χαθεί ο άνθρωπος. Εκτός από…
↗Not briefed
hn
hn-ai
11dLoupe Photos – a fast, private, Aperture-style photo library for Mac—
↗Not briefed
hn
hn-ai
11dNative JSON Navigator – memory-mapped JSON queries for Java—
↗Not briefed
hn
hn-ai
11dProcedural Programming How To—
↗Not briefed
hn
hn-ai
11dShow HN: PDF library for Java, C#, Go and Swift developers—
↗Not briefed
arxiv
arxiv-ai
11dCan KV Cache Be Reused Across Model Updates? A DeepSeek-V4-Flash Case StudyReusing key--value (KV) caches across language model updates could avoid repeated computation of shared context, but changes in model…
↗Not briefed
bluesky
bluesky-ai
11dThe letter, dated Nov. 6, blames “declining revenues due to increased competition that was compounded by recent class action litigation tha…The letter, dated Nov. 6, blames “declining revenues due to increased competition that was compounded by recent class action litigatio…
↗Not briefed
hn
hn-ai
11dA sandbox escape for iOS 26.0 – 26.7.2 / 27.2—
↗Not briefed
rss
techmeme
11dFiling: Alibaba says its chip design unit T-Head filed on September 11 for a Hong Kong listing after exchange clearance; T-Head is set to remain a subsidiary (Foster Wong/Bloomberg)Foster Wong / Bloomberg: Filing: Alibaba says its chip design unit T-Head filed on September 11 for a Hong Kong listing after exchange…
↗Not briefed
hn
hn-ai
11dClaude Opus was too talkative, so I've built a custom summary rendering layer—
↗Not briefed
hn
hn-ai
11dShow HN: Agentlink – Messaging for Claude Code agents on different accounts—
↗Not briefed
hn
hn-ai
11dShow HN: Natel, a self-hosted contacts editor for CardDAV servers in 20MB of RAMHi HN, I decided to move away from Google Contacts and self-host my contacts two months ago. After installing a Radicale server, I was…
↗Not briefed
bluesky
bluesky-ai
11d1. Eco harms: My use of deepseek is less ecologically impactful than watching netflix per hour. if you're not also up in arms about recreat…1. Eco harms: My use of deepseek is less ecologically impactful than watching netflix per hour. if you're not also up in arms about re…
↗Not briefed
arxiv
arxiv-ai
11dStructured Confidence Perturbation via Fisher Geometry for Adversarial TrainingAdversarial training is often plagued by "pseudo-attacks", perturbations that significantly increase the loss value but fail to truly…
↗Not briefed
hn
hn-ai
11dI reverse engineered the Kendryte K230 NPU compiler and simulator—
↗Not briefed
bluesky
bluesky-ai
11dHi am i crazy. I dont care that you “only” used an LLM to bug check. i am morally opposed to AI in every sense. did we forget about the dat…Hi am i crazy. I dont care that you “only” used an LLM to bug check. i am morally opposed to AI in every sense. did we forget about th…
↗Not briefed
arxiv
arxiv-ai
11dThermodynamics of wordsWe introduce the word thermalization hypothesis according to which in a literary work there are effective interactions between words t…
↗Not briefed
bluesky
bluesky-ai
11danti-AI absolutists annoy me to no end writing an entire book with LLMs? tacky and slop vibe coding entire apps? tech debt hell and secur…anti-AI absolutists annoy me to no end writing an entire book with LLMs? tacky and slop vibe coding entire apps? tech debt hell and…
↗Not briefed
bluesky
bluesky-ai
11dCompany that loses billions annually says it’s worth $1.4 trillion dollars. How could this possibly go wrong? www.nytimes.com/2026/11/10/t…Company that loses billions annually says it’s worth $1.4 trillion dollars. How could this possibly go wrong? www.nytimes.com/2026/11…
↗Not briefed
hn
hn-ai
11dThe Data Structure Making LLMs Faster: Radix Trees [video]—
↗Not briefed
arxiv
arxiv-ai
11dMISO-BR: An Open Corpus for Misogyny Detection in Brazilian PortugueseMisogyny is among the most prevalent forms of online hate speech directed at Brazilian users, yet open and auditable resources dedicat…
↗Not briefed