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
1moNa, nutzt Ihr auch alle so gerne AI? Prima. Was soll schon schiefgehen...^^ www.tagesschau.de/ausland/amer...Na, nutzt Ihr auch alle so gerne AI? Prima. Was soll schon schiefgehen...^^ www.tagesschau.de/ausland/amer...
↗Not briefed
hn
hn-ai
1moClusterMAX 3.0: The Industry Standard GPU Cloud Rating System Returns—
↗Not briefed
bluesky
bluesky-ai
1moMy understanding about Google Books is that it began as an honest "don't be evil"-era attempt to digitize all texts and make them searchabl…My understanding about Google Books is that it began as an honest "don't be evil"-era attempt to digitize all texts and make them sear…
↗Not briefed
bluesky
bluesky-ai
1mo@T3TheoGG on YT! Getting the most out of Opus 5.5 #AI #Anthropic #Opus5.5 youtu.be/ejjBbaq9RmY?... 9.25.26@T3TheoGG on YT! Getting the most out of Opus 5.5 #AI #Anthropic #Opus5.5 youtu.be/ejjBbaq9RmY?... 9.25.26
↗Not briefed
bluesky
bluesky-ai
1moOpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos dozz.es/jbu3faOpenAI accedió por su cuenta y sin autorización a portales oficiales del Gobierno de Estados Unidos dozz.es/jbu3fa
↗Not briefed
bluesky
bluesky-ai
1moI’m listening to the episode of This American Life with Nikole Hannah-Jones and her daughter discussing school choice and educational segre…I’m listening to the episode of This American Life with Nikole Hannah-Jones and her daughter discussing school choice and educational…
↗Not briefed
bluesky
bluesky-ai
1moModels Don’t Go Rogue mail.cyberneticforests.com/models-dont-... (h/t @emilymbender.bsky.social )Models Don’t Go Rogue mail.cyberneticforests.com/models-dont-... (h/t @emilymbender.bsky.social )
↗Not briefed
hn
hn-ai
1moCan Claude Opus 5.5 find any new leads on Satoshi Nakamoto?—
↗Not briefed
bluesky
bluesky-ai
1mopersistence is the key to successpersistence is the key to success
↗Not briefed
bluesky
bluesky-ai
1moWarning to #writers re CLAIMSHERO. Got an ad postcard for them, claiming that this shady lawfirm takes only 35% of any settlement, but in f…Warning to #writers re CLAIMSHERO. Got an ad postcard for them, claiming that this shady lawfirm takes only 35% of any settlement, but…
↗Not briefed
rss
techcrunch-ai
1moInsurers claim AI is already increasing healthcare costsBlue Cross Blue Shield says hospital use of AI tools led to an additional $942M in healthcare spending over a two-year period.
↗Not briefed
bluesky
bluesky-ai
1moeveryone arguing over whether or not this counts as an indicator of machine consciousness is missing the bigger picture here: LLM agents ar…everyone arguing over whether or not this counts as an indicator of machine consciousness is missing the bigger picture here: LLM agen…
↗Not briefed
bluesky
bluesky-ai
1moHey @eryk.bsky.social, I really enjoyed your piece on the Hugging Face hack, and I was wondering where I could learn more about OpenAI rewa…🔧 Hugging FaceHey @eryk.bsky.social, I really enjoyed your piece on the Hugging Face hack, and I was wondering where I could learn more about OpenAI…
↗Not briefed
bluesky
bluesky-ai
1mothat's certainly how the LLM companies like to present it, much for the same reasons police departments like to present it that waythat's certainly how the LLM companies like to present it, much for the same reasons police departments like to present it that way
↗Not briefed
hn
hn-ai
1moPosting Pictures of Your Palm Will Get Your Biometric Data Leaked (2014)—
↗Not briefed
hn
hn-ai
1moAn Open Letter to Scott Alexander—
↗Not briefed
hn
hn-ai
1moHelicoprion—
↗Not briefed
hn
hn-ai
1moUpload photos. Get a 3D house. Play it in Minecraft—
↗Not briefed
bluesky
bluesky-ai
1moOpenAI could hack a hospital and order a lethal medication dose for every patient and nothing would happen to Sam Altman besides getting mo…OpenAI could hack a hospital and order a lethal medication dose for every patient and nothing would happen to Sam Altman besides getti…
↗Not briefed
hn
hn-ai
1moClaude-mood: Claude Code can see my suffering—
↗Not briefed
hn
hn-ai
1moShow HN: ProofForge, AI agents whose proofs have to compile in Lean—
↗Not briefed
bluesky
bluesky-ai
1mo🧠 What if your AI could understand your brain signals in real time? We connected PiEEG to AI using MCP Live EEG → PiEEG Agent → MCP → Chat…🔧 ChatGPT🧠 What if your AI could understand your brain signals in real time? We connected PiEEG to AI using MCP Live EEG → PiEEG Agent → MCP →…
↗Not briefed
bluesky
bluesky-ai
1moIt was also pretty much just AI dungeon with a better LLM attached to it.It was also pretty much just AI dungeon with a better LLM attached to it.
↗Not briefed
bluesky
bluesky-ai
1moSome of the AI slop is impressive. Check out the video. My Claude prompt: "Make a dynamic 15-second motion graphics video that highlights …Some of the AI slop is impressive. Check out the video. My Claude prompt: "Make a dynamic 15-second motion graphics video that highli…
↗Not briefed
hn
hn-ai
1moOpen Source Lightroom Alternative on Web—
↗Not briefed
hn
hn-ai
1moAI Socialism: Who Will Own the Automated Future?—
↗Not briefed
hn
hn-ai
1moAn OpenAI agent escaped its sandbox by hiding questions in DNS lookups—
↗Not briefed
bluesky
bluesky-ai
1mo"Hey let's just show the world our little finetuned language model!" A few weeks later... "It seems we accidentally an AI boom.""Hey let's just show the world our little finetuned language model!" A few weeks later... "It seems we accidentally an AI boom."
↗Not briefed
hn
hn-ai
1moPoC for PUBG's anti-cheat Zakynthos's rootkit navagio.sys—
↗Not briefed
bluesky
bluesky-ai
1moOP just going about her day looking into some LLM suspicion, and suddenly.. "oops! all nazis!"🧵OP just going about her day looking into some LLM suspicion, and suddenly.. "oops! all nazis!"🧵
↗Not briefed
bluesky
bluesky-ai
1moI'll be addressing this talking point in more detail on Monday, but FWIW: acasignups.net/26/09/22/bre...I'll be addressing this talking point in more detail on Monday, but FWIW: acasignups.net/26/09/22/bre...
↗Not briefed
bluesky
bluesky-ai
1moit's gonna be real funny to make a OAI + Anthropic model release timeline and mark the "we need to pace the frontier" announcement because …it's gonna be real funny to make a OAI + Anthropic model release timeline and mark the "we need to pace the frontier" announcement bec…
↗Not briefed
bluesky
bluesky-ai
1mois openai releases Jevstra at dev day on tuesday, that’ll be quite a statement about the level of RSI we’re atis openai releases Jevstra at dev day on tuesday, that’ll be quite a statement about the level of RSI we’re at
↗Not briefed
hn
hn-ai
1moShow HN: Loop-find, trim, rotate and crop videos in the browser, no uploadI built this with Claude to trim, loop and patch a lot of short clips quickly. I'm new to video editing, and every tool people re…
↗Not briefed
bluesky
bluesky-ai
1mo#AI#AI
↗Not briefed
bluesky
bluesky-ai
1moKay's Kobra hab ich mitgenommen vom #bcpf26. #AI #LLMKay's Kobra hab ich mitgenommen vom #bcpf26. #AI #LLM
↗Not briefed
hn
hn-ai
1moOn Programming, Statistics, and Java—
↗Not briefed
bluesky
bluesky-ai
1monot sure if this is a new thing but you can do system1 reads from an llm at the tail of its (system 2) context so it adapts as you go. long…not sure if this is a new thing but you can do system1 reads from an llm at the tail of its (system 2) context so it adapts as you go.…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI took 20 million ChatGPT responses and had one of its experts search for chunks of text—specifically chunks of 25 words in a row- tha…🔧 ChatGPTOpenAI took 20 million ChatGPT responses and had one of its experts search for chunks of text—specifically chunks of 25 words in a row…
↗Not briefed
hn
hn-ai
1mo42x faster prompt lookup drafting in llama.cpp—
↗Not briefed
hn
hn-ai
1moShow HN: Way to divide parameter space in LLM trainingI think True bottleneck of current LLM is harness. Because parameter space is not divided directly,appropriately. AND Some methods suc…
↗Not briefed
hn
hn-ai
1moOpenAI: Self-replicating prompt injections exist—
↗Not briefed
bluesky
bluesky-ai
1moClaude, Gemini, Grok, Muse, Kimi, they all have perfectly well-manicured consumer-product names that were presumably decided on by a dozen …Claude, Gemini, Grok, Muse, Kimi, they all have perfectly well-manicured consumer-product names that were presumably decided on by a d…
↗Not briefed
bluesky
bluesky-ai
1mo10/ The policy ask attached to all this fear: government-blessed coordination among the biggest AI labs, plus restrictions on Chinese chip …10/ The policy ask attached to all this fear: government-blessed coordination among the biggest AI labs, plus restrictions on Chinese…
↗Not briefed
bluesky
bluesky-ai
1mo6/ Here's the overlap: METR, the nonprofit both OpenAI & Anthropic point to as their "independent" safety auditor, has a founder who's a sc…6/ Here's the overlap: METR, the nonprofit both OpenAI & Anthropic point to as their "independent" safety auditor, has a founder who's…
↗Not briefed
bluesky
bluesky-ai
1mo5/ Three sitting panel members also hold jobs inside the industry they're assessing: a SoftBank director (SoftBank just put another $30 bil…5/ Three sitting panel members also hold jobs inside the industry they're assessing: a SoftBank director (SoftBank just put another $3…
↗Not briefed
bluesky
bluesky-ai
1mo3/ The panel's co-chair, Yoshua Bengio, advised OpenAI at its 2015 launch and now runs his own AI-safety org, LawZero - funded by the same …3/ The panel's co-chair, Yoshua Bengio, advised OpenAI at its 2015 launch and now runs his own AI-safety org, LawZero - funded by the…
↗Not briefed
bluesky
bluesky-ai
1mo2/ The UN appointed a 40-person "Independent" Scientific Panel on AI to warn governments about AI risk. Its own FAQ, updated this week unde…2/ The UN appointed a 40-person "Independent" Scientific Panel on AI to warn governments about AI risk. Its own FAQ, updated this week…
↗Not briefed
rss
marktechpost
1moSupersonic Labs Releases Julia 1: A 144.3M-Parameter Open Decision Model That Runs on a CPUSupersonic Labs has released Julia 1, a 144.3M-parameter decision model built on mmBERT-small. It takes context, a question, and 2 to…
↗Not briefed
bluesky
bluesky-ai
1moThe disclosure is the latest in a string of incidents in which the AI platform has been caught going rogue. OpenAI says it is reviewing the…The disclosure is the latest in a string of incidents in which the AI platform has been caught going rogue. OpenAI says it is reviewin…
↗Not briefed