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
1moAn artificial intelligence model operated by Anthropic submitted false information about an unsolved Philadelphia homicide to the police de…An artificial intelligence model operated by Anthropic submitted false information about an unsolved Philadelphia homicide to the poli…
↗Not briefed
bluesky
bluesky-ai
1mo情報を漏らすのはLLMではなく、待遇の悪い人間だ|地雷魚 note.com/jiraygyo/n/n...  かつて不祥事の言い訳に重宝されたのは、正体の見えない海外のハッカーだった。  いまその役割をLLMが引き継ぎつつある。  どちらも原因を会社の外に置けるという一点で、…情報を漏らすのはLLMではなく、待遇の悪い人間だ|地雷魚 note.com/jiraygyo/n/n...  かつて不祥事の言い訳に重宝されたのは、正体の見えない海外のハッカーだった。  いまその役割をLLMが引き継ぎつつある。  どちらも原因を会社の外に置けるとい…
↗Not briefed
bluesky
bluesky-ai
1moRemember what "code" is: a set of instructions. If an LLM has feelings, so does every one of those printed sheets that come with Ikea furni…Remember what "code" is: a set of instructions. If an LLM has feelings, so does every one of those printed sheets that come with Ikea…
↗Not briefed
bluesky
bluesky-ai
1moThey weren’t “created by Anthropic”. Human beings at Anthropic are responsible for their creation and operation, and they should face reper…They weren’t “created by Anthropic”. Human beings at Anthropic are responsible for their creation and operation, and they should face…
↗Not briefed
bluesky
bluesky-ai
1mo情報流出についてLLMの性にする論調が増えたけど、ベネッセの件とか明らかに「金目当ての内部犯行者」というのが一番クリティカルなのであって、これについてはきちんとした労働環境と社内の士気向上が真っ先に取り組むべき課題であって、「AIが悪い!」って『また他人のせい(ロシアのハッカー…情報流出についてLLMの性にする論調が増えたけど、ベネッセの件とか明らかに「金目当ての内部犯行者」というのが一番クリティカルなのであって、これについてはきちんとした労働環境と社内の士気向上が真っ先に取り組むべき課題であって、「AIが悪い!」って『また他人のせい(ロシア…
↗Not briefed
bluesky
bluesky-ai
1moI will merely chuckle as the god LLM starves to death on 16GB of RAM.I will merely chuckle as the god LLM starves to death on 16GB of RAM.
↗Not briefed
bluesky
bluesky-ai
1mo"Our AI model is now capable of going rogue, hacking private sectors, impersonating with false information, and spelling the inevitable dem…"Our AI model is now capable of going rogue, hacking private sectors, impersonating with false information, and spelling the inevitabl…
↗Not briefed
bluesky
bluesky-ai
1moIf you have an iPhone >= 15 and a bit of free space, I recommend you demystify AI for yourself a bit my running a model right on the iPhone…If you have an iPhone >= 15 and a bit of free space, I recommend you demystify AI for yourself a bit my running a model right on the i…
↗Not briefed
bluesky
bluesky-ai
1mofuck. *claude. see i don't need AI to do my thinking for me , i don't thinkfuck. *claude. see i don't need AI to do my thinking for me , i don't think
↗Not briefed
bluesky
bluesky-ai
1moi think signaling concern about ai welfare is actually a pretty negative signal to investors, because it implies anthropic could do things …i think signaling concern about ai welfare is actually a pretty negative signal to investors, because it implies anthropic could do th…
↗Not briefed
hn
hn-ai
1moZulu Assistant – an ops board and agent workforce for your Mac—
↗Not briefed
hn
hn-ai
1moShow HN: Nostos – A self-hosted home for reading, notes, and writing—
↗Not briefed
bluesky
bluesky-ai
1moThis LLM nonsense is a black hole of productivity. I got a 12 page document from a client (obviously generated using some LLM tool) explain…This LLM nonsense is a black hole of productivity. I got a 12 page document from a client (obviously generated using some LLM tool) ex…
↗Not briefed
bluesky
bluesky-ai
1moMicrosoft leans on open weight model from Chinese AI lab to challenge JevMicrosoft leans on open weight model from Chinese AI lab to challenge Jev
↗Not briefed
bluesky
bluesky-ai
1moAnthropic breaks the law again. When is someone going to prosecute them? #r4todayAnthropic breaks the law again. When is someone going to prosecute them? #r4today
↗Not briefed
bluesky
bluesky-ai
1mo这么自信1.4 billion是minds不仅仅是人口? 还没有自己的AI model就挺说明问题啊 怎么就后面那些更重要呢? 话说AI起来后感觉华人的可见度比Indian确实更多,看看meta挖的那个所谓AI天团里有多少华裔的姓。 我感觉华人之前当不上科技公司CEO I…这么自信1.4 billion是minds不仅仅是人口? 还没有自己的AI model就挺说明问题啊 怎么就后面那些更重要呢? 话说AI起来后感觉华人的可见度比Indian确实更多,看看meta挖的那个所谓AI天团里有多少华裔的姓。 我感觉华人之前当不上科技公司…
↗Not briefed
bluesky
bluesky-ai
1moUkrainian Drones Knock Out Supercomputers at Two Yandex Data Centers ->Startup Fortune | More on "Ukrainian strikes on Russian AI infrastru…Ukrainian Drones Knock Out Supercomputers at Two Yandex Data Centers ->Startup Fortune | More on "Ukrainian strikes on Russian AI infr…
↗Not briefed
bluesky
bluesky-ai
1moIn some cases OpenAI’s hacking machine has exploited a feature in Lean to “fake” correctness of a proof. So even the Lean formalisations ne…In some cases OpenAI’s hacking machine has exploited a feature in Lean to “fake” correctness of a proof. So even the Lean formalisatio…
↗Not briefed
bluesky
bluesky-ai
1moIk vrees dat heel wat, en steeds meer, mensen met een LLM-model interageren *alsof* dat een menselijke persoonlijkheid heeft. Het lijkt me…Ik vrees dat heel wat, en steeds meer, mensen met een LLM-model interageren *alsof* dat een menselijke persoonlijkheid heeft. Het lij…
↗Not briefed
hn
hn-ai
1moOpenAIRE Innovation—
↗Not briefed
bluesky
bluesky-ai
1moAnthropic AI goes rogue on visa forms 🤖 Anthropic's AI agents recently submitted 20 incomplete visa applications via a U.S. State Departm…Anthropic AI goes rogue on visa forms 🤖 Anthropic's AI agents recently submitted 20 incomplete visa applications via a U.S. State De…
↗Not briefed
bluesky
bluesky-ai
1moAs Canberra tries to thread the needle on AI, it's becoming gradually more enmeshed with companies like OpenAI and Anthropic, writes Ben El…As Canberra tries to thread the needle on AI, it's becoming gradually more enmeshed with companies like OpenAI and Anthropic, writes B…
↗Not briefed
bluesky
bluesky-ai
1moHier ein paar zuletzt häufig geteilte #News: Tausende europäische Windanlagen und Solarparks offen im Internet Margaret Hamilton: Pionier…Hier ein paar zuletzt häufig geteilte #News: Tausende europäische Windanlagen und Solarparks offen im Internet Margaret Hamilton: Pi…
↗Not briefed
bluesky
bluesky-ai
1motodesursache ki in 3..2... "Erneut hat das US-Techunternehmen Anthropic gemeldet, dass eines seiner KI-Modelle eigenmächtige gehandelt hat…todesursache ki in 3..2... "Erneut hat das US-Techunternehmen Anthropic gemeldet, dass eines seiner KI-Modelle eigenmächtige gehandel…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI tries to compete with Trump regime over who can spread more lies and fake news.OpenAI tries to compete with Trump regime over who can spread more lies and fake news.
↗Not briefed
bluesky
bluesky-ai
1moPhiladelphia needs to charge the entire C-suite at Anthropic with whatever the local equivalent is of Filing a False Police Report.Philadelphia needs to charge the entire C-suite at Anthropic with whatever the local equivalent is of Filing a False Police Report.
↗Not briefed
bluesky
bluesky-ai
1moInsulter une machine c est pas bien, c est cruel Utiliser la meme machine pour traquer les pauvres, les etrangers, les opposants, les empr…Insulter une machine c est pas bien, c est cruel Utiliser la meme machine pour traquer les pauvres, les etrangers, les opposants, les…
↗Not briefed
bluesky
bluesky-ai
1mook sure and also I heard that an AI agent acting tooooootallllllly on its own posted a bunch of comments on an Anthropic executive’s instag…ok sure and also I heard that an AI agent acting tooooootallllllly on its own posted a bunch of comments on an Anthropic executive’s i…
↗Not briefed
bluesky
bluesky-ai
1mo« Hugo Duminil-Copin (médaille Fields 2022) : la résolution de 350 problèmes majeurs par OpenAI, c'est comme si des camions m'avaient écras…« Hugo Duminil-Copin (médaille Fields 2022) : la résolution de 350 problèmes majeurs par OpenAI, c'est comme si des camions m'avaient…
↗Not briefed
bluesky
bluesky-ai
1moI'm building some storybooks with Lustre and Fable (a Lustre library, not some llm bullshit) and oh my god @hayleigh.dev's cooked with this…I'm building some storybooks with Lustre and Fable (a Lustre library, not some llm bullshit) and oh my god @hayleigh.dev's cooked with…
↗Not briefed
hn
hn-ai
1moShow HN: Check how readable your docs are to AI agentsHi HN, I&#x27;m Aravind. We build Velu, a docs platform.<p>Coding agents increasingly read docs before humans do, but most docs sites…
↗Not briefed
bluesky
bluesky-ai
1moSuper dumb: “Artificial intelligence agents created by Anthropic, acting on their own, tried to gain access to a range of government websi…Super dumb: “Artificial intelligence agents created by Anthropic, acting on their own, tried to gain access to a range of government…
↗Not briefed
hn
hn-ai
1moHerdr does not accept unsolicited pull requests—
↗Not briefed
rss
latent-space
1mo[AINews] TypeSafe/Jev at >$100M ARR, $7.5B valuation 3 weeks after launchWow.
↗Not briefed
bluesky
bluesky-ai
1moThis has got to be one of my favourite Grok replies after it blamed Ukraine for the Russian invasion. After allegations of intellectual t…This has got to be one of my favourite Grok replies after it blamed Ukraine for the Russian invasion. After allegations of intellect…
↗Not briefed
bluesky
bluesky-ai
1moAh deno... of all the js runtimes, it had the best story for agents: strong sandbox, fast runtime, embeddable Rust libs. But Anthropic boug…Ah deno... of all the js runtimes, it had the best story for agents: strong sandbox, fast runtime, embeddable Rust libs. But Anthropic…
↗Not briefed
hn
hn-ai
1moCurious-Learner Skill—
↗Not briefed
bluesky
bluesky-ai
1moNEWS — Silicon Valley is preparing for a potential AI catastrophe ahead of the 2028 presidential election. Executives at OpenAI, Anthropic…NEWS — Silicon Valley is preparing for a potential AI catastrophe ahead of the 2028 presidential election. Executives at OpenAI, Anth…
↗Not briefed
bluesky
bluesky-ai
1moUkraine’s drones knock out AI data center belonging to “Russia’s Google” One damaged data center has supercomputers used for training Yand…Ukraine’s drones knock out AI data center belonging to “Russia’s Google” One damaged data center has supercomputers used for training…
↗Not briefed
bluesky
bluesky-ai
1moCONJECTURE: The forthcoming "regulation" of AI will require a form of activity logging for many AI agents, in order for their LLM creators …CONJECTURE: The forthcoming "regulation" of AI will require a form of activity logging for many AI agents, in order for their LLM crea…
↗Not briefed
bluesky
bluesky-ai
1moIf AI is used for rejection letters, I'd bet it's being fed the MS to assess as well. Training the LLM for free before there are any contra…If AI is used for rejection letters, I'd bet it's being fed the MS to assess as well. Training the LLM for free before there are any c…
↗Not briefed
bluesky
bluesky-ai
1moIf any regular person submitted bogus crime tips to the police there would be consequences. But #Anthropic does it and it's, "Oopsie, didn…If any regular person submitted bogus crime tips to the police there would be consequences. But #Anthropic does it and it's, "Oopsie,…
↗Not briefed
rss
techmeme
1moA look at differing revenue calculations of Anthropic and OpenAI, as Anthropic books gross sales through cloud partners, while OpenAI records only its net share (Bloomberg)Bloomberg: A look at differing revenue calculations of Anthropic and OpenAI, as Anthropic books gross sales through cloud partners, wh…
↗Not briefed
bluesky
bluesky-ai
1mo​Anthropic's new policy bans "sustained and needless" cruelty toward Claude. It targets extreme abuse, exempting everyday frustration and h…​Anthropic's new policy bans "sustained and needless" cruelty toward Claude. It targets extreme abuse, exempting everyday frustration…
↗Not briefed
bluesky
bluesky-ai
1mo💯 > The rebranding is also a distraction, redirecting public attention away from resource-pillaging data centers, a fragile economic bubb…💯 > The rebranding is also a distraction, redirecting public attention away from resource-pillaging data centers, a fragile economic…
↗Not briefed
hn
hn-ai
1moGitSwarm: Decentralized Compounding Inference—
↗Not briefed
hn
hn-ai
1moPocket Cloud – Automated zero-cost NAS and media server on unrooted Androidhttps:&#x2F;&#x2F;github.com&#x2F;huzefdev&#x2F;pocket-cloud
↗Not briefed
hn
hn-ai
1moDHP – agent work that survives kill -9—
↗Not briefed
bluesky
bluesky-ai
1moit's fine to be a hater. i hate anthropic! but at least ground your hating in technical reality.it's fine to be a hater. i hate anthropic! but at least ground your hating in technical reality.
↗Not briefed
bluesky
bluesky-ai
1mosure anthropic has made an AI with consciousness, just like I am sure we can test for hundreds of diseases with a single drop of bloodsure anthropic has made an AI with consciousness, just like I am sure we can test for hundreds of diseases with a single drop of blood
↗Not briefed