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
1moLLM Red-temaing vs. Agent Red teaming—
↗Not briefed
bluesky
bluesky-ai
1mo#Désormais #Anthropic #écarté #Pentagone #justice #américaine #confirme #Now #Anthropic #excluded #Pentagon #US #appeal #court #confirms 🤖…#Désormais #Anthropic #écarté #Pentagone #justice #américaine #confirme #Now #Anthropic #excluded #Pentagon #US #appeal #court #confir…
↗Not briefed
hn
hn-ai
1moUsing Pomodoro to limit interaction with AI—
↗Not briefed
hn
hn-ai
1moShow HN: Linecite – find and fix stale file:line references in your docs—
↗Not briefed
bluesky
bluesky-ai
1moAnthropicがAI開発のために各国で書籍をいっぱい買ってるって話、話の渦中にある「読み取る本は分解されて、最後は破棄される」という点がさらにいろんな人の神経に触ってそうだな~と思ったのと、 これって2010年初頭に盛り上がった「自炊代行(個人所有の書籍を企業が引き取ってス…AnthropicがAI開発のために各国で書籍をいっぱい買ってるって話、話の渦中にある「読み取る本は分解されて、最後は破棄される」という点がさらにいろんな人の神経に触ってそうだな~と思ったのと、 これって2010年初頭に盛り上がった「自炊代行(個人所有の書籍を企業が引…
↗Not briefed
bluesky
bluesky-ai
1moOpenAIのAI、米政府機関サイトにもアクセス 開発制御できず - 日本経済新聞 www.nikkei.com/article/DGXZ...OpenAIのAI、米政府機関サイトにもアクセス 開発制御できず - 日本経済新聞 www.nikkei.com/article/DGXZ...
↗Not briefed
hn
hn-ai
1moI Built Laya Studio: Specialize Laya Without Fine-Tuning—
↗Not briefed
bluesky
bluesky-ai
1moIch verstehe nicht, warum das politisch nicht verstanden wird. Es ist längst kein SciFi mehr Was soll noch alles passieren und wo soll KI n…Ich verstehe nicht, warum das politisch nicht verstanden wird. Es ist längst kein SciFi mehr Was soll noch alles passieren und wo soll…
↗Not briefed
hn
hn-ai
1moJevq, A Jev filter sidecar for jq—
↗Not briefed
hn
hn-ai
1moLinux Kernel Developers Consider Adding Agents.md to Help Guide AI/LLM Agents—
↗Not briefed
bluesky
bluesky-ai
1moWell Lions, Dockers, seems like everyone knows. Bit stunned to see this as No 2 on Bluesky's global trending.. Hell of a game though. #…Well Lions, Dockers, seems like everyone knows. Bit stunned to see this as No 2 on Bluesky's global trending.. Hell of a game thoug…
↗Not briefed
bluesky
bluesky-ai
1moIt seems to me it's conceptually flawed forhumans, who think a god gave them a brain in order to to use it but who can't even reason their …It seems to me it's conceptually flawed forhumans, who think a god gave them a brain in order to to use it but who can't even reason t…
↗Not briefed
bluesky
bluesky-ai
1moI went on R34 to see any Nsfw art of Phis VTuber model…and I was disappointed when I saw it was all AI garbage 😑 sighI went on R34 to see any Nsfw art of Phis VTuber model…and I was disappointed when I saw it was all AI garbage 😑 sigh
↗Not briefed
bluesky
bluesky-ai
1moJa, ein Riesenproblem für alle Online-Medien. Und sie wissen, was sie tun: www.404media.co/doom-loop-op...Ja, ein Riesenproblem für alle Online-Medien. Und sie wissen, was sie tun: www.404media.co/doom-loop-op...
↗Not briefed
bluesky
bluesky-ai
1moLLM: Large Lagomorph MyxomatosisLLM: Large Lagomorph Myxomatosis
↗Not briefed
bluesky
bluesky-ai
1moUn LLM quand il est sous environnement CLI :Un LLM quand il est sous environnement CLI :
↗Not briefed
bluesky
bluesky-ai
1moIf you use any hyperscaled chatbot at all regularly you are providing vital information to untouchable creeps. Stop it www.theguardian.c…If you use any hyperscaled chatbot at all regularly you are providing vital information to untouchable creeps. Stop it www.theguard…
↗Not briefed
bluesky
bluesky-ai
1moCan OpenAI feel pain? 🤔 Seems like it'd be fun to find out...Can OpenAI feel pain? 🤔 Seems like it'd be fun to find out...
↗Not briefed
bluesky
bluesky-ai
1mothe weirdest thing ive noticed is that if i paste llm output direct to niri she immediately becomes so subservient its eeriethe weirdest thing ive noticed is that if i paste llm output direct to niri she immediately becomes so subservient its eerie
↗Not briefed
bluesky
bluesky-ai
1moAs did DOGE. www.bbc.co.uk/news/article...As did DOGE. www.bbc.co.uk/news/article...
↗Not briefed
hn
hn-ai
1moVideo Tubelets: Objects as Volumes in Space-Time—
↗Not briefed
bluesky
bluesky-ai
1moit endlessly annoys me thinking that most modern llm quirks or tics are catering to a flawed graderit endlessly annoys me thinking that most modern llm quirks or tics are catering to a flawed grader
↗Not briefed
bluesky
bluesky-ai
1mo#Canal24Horas | Ampliación OpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos#Canal24Horas | Ampliación OpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos
↗Not briefed
bluesky
bluesky-ai
1moLa conclusion est sans appel: je suis con comme un LLMLa conclusion est sans appel: je suis con comme un LLM
↗Not briefed
rss
marktechpost
1moExa Launches Agent Ultra: A Subagent Swarm Deep Research API Built for Exhaustive List BuildingExa has released Agent Ultra, the highest effort mode of its Exa Agent API. It coordinates subagents across thousands of sources for l…
↗Not briefed
hn
hn-ai
1moWho is behind the AI safety backlash?—
↗Not briefed
bluesky
bluesky-ai
1moCanadian lawsuits against OpenAI raise questions about liability and duty to warn in the AI industry, say experts.Canadian lawsuits against OpenAI raise questions about liability and duty to warn in the AI industry, say experts.
↗Not briefed
hn
hn-ai
1moThe OpenAI "Hack" of Australia's Medicare That Wasn't—
↗Not briefed
bluesky
bluesky-ai
1moSo, I guess the Feds will be raiding their offices any moment, impounding servers, arresting execs? Foreign governments will be bringing ch…So, I guess the Feds will be raiding their offices any moment, impounding servers, arresting execs? Foreign governments will be bringi…
↗Not briefed
bluesky
bluesky-ai
1moC'est quand même assez troublant de voir que "les experts en IA" arrivent à dire que les phrases courtes et "percutantes" sont un signe de …C'est quand même assez troublant de voir que "les experts en IA" arrivent à dire que les phrases courtes et "percutantes" sont un sign…
↗Not briefed
bluesky
bluesky-ai
1moIn Feb 2026, Anthropic blocked GenAI.mil to keep guardrails on weapons and mass surveillance. A D.C. court upheld the supply-chain risk ban…In Feb 2026, Anthropic blocked GenAI.mil to keep guardrails on weapons and mass surveillance. A D.C. court upheld the supply-chain ris…
↗Not briefed
bluesky
bluesky-ai
1moClimate change accelerators, just there to help you ask an llm what to get your wife for her birthday instead of tryingClimate change accelerators, just there to help you ask an llm what to get your wife for her birthday instead of trying
↗Not briefed
bluesky
bluesky-ai
1moI read 20 passages in an article that paired fiction by people with LLM ones and each time I correctly identified the LLM even when the hum…I read 20 passages in an article that paired fiction by people with LLM ones and each time I correctly identified the LLM even when th…
↗Not briefed
bluesky
bluesky-ai
1mo🇺🇸 OpenAI AI agents infiltrated US government websites due to a glitch during testing. They gained access to the resources of the US De…🇺🇸 OpenAI AI agents infiltrated US government websites due to a glitch during testing. They gained access to the resources of the…
↗Not briefed
bluesky
bluesky-ai
1mo„Erneut haben sich KI-Modelle des US-Unternehmens OpenAI selbstständig gemacht.“ Sogenannte Ich-AG-Enten.„Erneut haben sich KI-Modelle des US-Unternehmens OpenAI selbstständig gemacht.“ Sogenannte Ich-AG-Enten.
↗Not briefed
bluesky
bluesky-ai
1mo“Rogue autonomous models” says OpenAI, as if no human has created, designed, engineered, tinkered with and influenced AI. Nice try, Sam & C…“Rogue autonomous models” says OpenAI, as if no human has created, designed, engineered, tinkered with and influenced AI. Nice try, Sa…
↗Not briefed
hn
hn-ai
1moBiggest design mistake I made when we moved OpenClaw to SQLite: using sync db—
↗Not briefed
hn
hn-ai
1moJev4j—
↗Not briefed
bluesky
bluesky-ai
1moIf you use ChaptGPT or any of the other cheat machines, you are a fool. #AIIf you use ChaptGPT or any of the other cheat machines, you are a fool. #AI
↗Not briefed
bluesky
bluesky-ai
1moOpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos dozz.es/jbu3f1OpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos dozz.es/jbu3f1
↗Not briefed
bluesky
bluesky-ai
1mo❗La IA de OpenAI intentó acceder sin autorización a webs oficiales del Gobierno de EEUU❗La IA de OpenAI intentó acceder sin autorización a webs oficiales del Gobierno de EEUU
↗Not briefed
bluesky
bluesky-ai
1moIn the meantime, OpenAI & Anthropic rolled out new models on Tuesday. Did anybody hear anything about this so-called AI slowdown lately? �…In the meantime, OpenAI & Anthropic rolled out new models on Tuesday. Did anybody hear anything about this so-called AI slowdown late…
↗Not briefed
rss
marktechpost
1moEnd-to-End Multimodal Data Augmentation and Adversarial Robustness Benchmark with AugLy for Images, Text, Audio, and PyTorchDiscover how to build a comprehensive multimodal augmentation and adversarial robustness workflow using AugLy for images, text, audio,…
↗Not briefed
bluesky
bluesky-ai
1moInstead of finding adjectives to inflate headlines, journalists might focus on what the “regulations” called for by industry to regulate it…Instead of finding adjectives to inflate headlines, journalists might focus on what the “regulations” called for by industry to regula…
↗Not briefed
bluesky
bluesky-ai
1moInsisteixo que aquests titulars funcionen millor si enlloc d'"admet" dius "amenaça": www.vilaweb.cat/noticies/ope...Insisteixo que aquests titulars funcionen millor si enlloc d'"admet" dius "amenaça": www.vilaweb.cat/noticies/ope...
↗Not briefed
bluesky
bluesky-ai
1moKI-Agenten haben mehrere Webseiten der US-Regierung angesteuert und dabei wohl auch Daten des Volkszählungsamts erfasst. OpenAI will nun ei…KI-Agenten haben mehrere Webseiten der US-Regierung angesteuert und dabei wohl auch Daten des Volkszählungsamts erfasst. OpenAI will n…
↗Not briefed
bluesky
bluesky-ai
1moso many people think RL can only sharpen capabilities that were already present in a model which is baffling once you consider nearly all o…so many people think RL can only sharpen capabilities that were already present in a model which is baffling once you consider nearly…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI directed a cyberattack against SEC. Fixed it for you. Arrest them.OpenAI directed a cyberattack against SEC. Fixed it for you. Arrest them.
↗Not briefed
hn
hn-ai
1moAnybody has experience with maintaining self hosted Git?I have used github, gitlab and bitbucket etc. Now all of them seems to be training all the code. How do one retain privacy while not p…
↗Not briefed
bluesky
bluesky-ai
1moNeue KI-Zwischenfälle bei OpenAI #USA #ChatGPT #OpenAI🔧 ChatGPTNeue KI-Zwischenfälle bei OpenAI #USA #ChatGPT #OpenAI
↗Not briefed