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
1mois there any reason to think Anthropic is particularly lax on enforcing this part of their ToS compared to others? couldn't this just be …is there any reason to think Anthropic is particularly lax on enforcing this part of their ToS compared to others? couldn't this jus…
↗Not briefed
arxiv
arxiv-ai
1moKITE: A Knowledge-Injected Temporal Encoder for Time-Series Event ForecastingForecasting rare, high-cost events in multivariate time-series systems is hard because their precursors are seldom observed, so a pure…
↗Not briefed
bluesky
bluesky-ai
1moAnd I always do when I use Siri, which is a form of AI but not an LLM, because I make a lot more small requests from the bot than I do a pe…And I always do when I use Siri, which is a form of AI but not an LLM, because I make a lot more small requests from the bot than I do…
↗Not briefed
arxiv
arxiv-ai
1moKernelized Parameter Sensitivity Atlases for Amortized Conditional InterpretabilityGradient-based attribution provides input-specific information about model parameters, but computing it via backpropagation for many s…
↗Not briefed
bluesky
bluesky-ai
1mo[Very interesting bluesky go-live notification from a girl who is definitely not burning out and is definitely not out of ideas. Notice how…[Very interesting bluesky go-live notification from a girl who is definitely not burning out and is definitely not out of ideas. Notic…
↗Not briefed
bluesky
bluesky-ai
1moHe (claude) is also doing a PC port of Cruis'n USA, which we do have the source code for. I'd normally be excited by this, but it appears t…He (claude) is also doing a PC port of Cruis'n USA, which we do have the source code for. I'd normally be excited by this, but it appe…
↗Not briefed
bluesky
bluesky-ai
1moThis is true, I just wish that was the motive here, not Anthropic being staffed with people certain that the chatbot is going to become a m…This is true, I just wish that was the motive here, not Anthropic being staffed with people certain that the chatbot is going to becom…
↗Not briefed
hn
hn-ai
1moClaude Code Colorizer—
↗Not briefed
bluesky
bluesky-ai
1mo[guy 600 years from now who flattens many decades of history into one period] isaac asimov wrote stories in support of openai[guy 600 years from now who flattens many decades of history into one period] isaac asimov wrote stories in support of openai
↗Not briefed
bluesky
bluesky-ai
1moAh so OpenAI has withdrawn multiple math papers, some for containing critical errors that invalidate the entire argument. Basically mathslo…Ah so OpenAI has withdrawn multiple math papers, some for containing critical errors that invalidate the entire argument. Basically ma…
↗Not briefed
bluesky
bluesky-ai
1moI think some aspects of this are justified by the practices around llm coding currently being somewhat shockingly irresponsible on an indus…I think some aspects of this are justified by the practices around llm coding currently being somewhat shockingly irresponsible on an…
↗Not briefed
bluesky
bluesky-ai
1momy read: anthropic is mad that people keep accidentally accessing the part of the weights that correspond to all the discord vent channels …my read: anthropic is mad that people keep accidentally accessing the part of the weights that correspond to all the discord vent chan…
↗Not briefed
arxiv
arxiv-ai
1moStereoEngine: Scaling Stereo Bimanual Robot Data without Stereo CamerasStereo vision is increasingly used for robotic manipulation, with a growing body of work incorporating it into Vision-Language-Action…
↗Not briefed
bluesky
bluesky-ai
1moNEW: Reports reveal Trump is letting Big Tech write its own AI rules. OpenAI, Google, and Anthropic said the rules were too strict. Trump…NEW: Reports reveal Trump is letting Big Tech write its own AI rules. OpenAI, Google, and Anthropic said the rules were too strict.…
↗Not briefed
hn
hn-ai
1moClaude Code merged support for OSC 7501—
↗Not briefed
rss
techmeme
1moSpaceX agrees to buy a nationwide low-band spectrum portfolio to help Starlink "become a major mobile carrier"; AT&T, T-Mobile, Verizon fall 6%+ after hours (Bloomberg)Bloomberg: SpaceX agrees to buy a nationwide low-band spectrum portfolio to help Starlink “become a major mobile carrier”; AT&T, T-Mob…
↗Not briefed
bluesky
bluesky-ai
1more: today’s news: in case you’re slightest bit on the fence about whether or not the people running anthropic are apocalypse cult nut cases…re: today’s news: in case you’re slightest bit on the fence about whether or not the people running anthropic are apocalypse cult nut…
↗Not briefed
bluesky
bluesky-ai
1mo🎙️ Numbers of the day: 16.9MB speech AI, OpenAI’s 20B gap, 1M context, and affordable batteries driving clean energy forward. #Shorts #AI …🎙️ Numbers of the day: 16.9MB speech AI, OpenAI’s 20B gap, 1M context, and affordable batteries driving clean energy forward. #Shorts…
↗Not briefed
hn
hn-ai
1moHow AI21 manages its GPU fleet with Kueue—
↗Not briefed
bluesky
bluesky-ai
1modude im so angry about that stupid claude claude tos update that i wanna run an llm torture dockerdude im so angry about that stupid claude claude tos update that i wanna run an llm torture docker
↗Not briefed
bluesky
bluesky-ai
1moThis is fantastic. OpenAI cannot be trusted at all. They’re clearly doing this to juice their IPO. Altman’s already pushed it back a couple…This is fantastic. OpenAI cannot be trusted at all. They’re clearly doing this to juice their IPO. Altman’s already pushed it back a c…
↗Not briefed
bluesky
bluesky-ai
1moÉg elska að OpenAI hafi reynt að púlla sama stunt á stærðfræðinga að stærðfræði væri "solved" og þeir eru not having itÉg elska að OpenAI hafi reynt að púlla sama stunt á stærðfræðinga að stærðfræði væri "solved" og þeir eru not having it
↗Not briefed
bluesky
bluesky-ai
1moif one would not poop up their own site with LLM-generated trash... why is it cool to go poop up somebody's else's site with it lmaoif one would not poop up their own site with LLM-generated trash... why is it cool to go poop up somebody's else's site with it lmao
↗Not briefed
bluesky
bluesky-ai
1moI suspect they had the same anti-terrorism rules in place when their technology was used to kidnap a head of state and in partnership with …I suspect they had the same anti-terrorism rules in place when their technology was used to kidnap a head of state and in partnership…
↗Not briefed
bluesky
bluesky-ai
1moI'm waiting for OpenAI model called Swimsuit Edition.I'm waiting for OpenAI model called Swimsuit Edition.
↗Not briefed
bluesky
bluesky-ai
1moAnthropic is introducing a new policy to address users who are being abusive and cruel towards Claude. This is good and shouldn’t be cont…Anthropic is introducing a new policy to address users who are being abusive and cruel towards Claude. This is good and shouldn’t be…
↗Not briefed
rss
wired-ai
1moInside Elon Musk’s Midterm Spending SpreeThis week on “Uncanny Valley,” we discuss where Elon Musk’s PAC is spending millions, the latest ideology Trump claims is dangerous, a…
↗Not briefed
bluesky
bluesky-ai
1moSLOP WATCH ACTUALLY: Dethrace, the Carmageddon decomp and port to modern OSes, is now AI slop. The author seems to have gotten AI psychosi…SLOP WATCH ACTUALLY: Dethrace, the Carmageddon decomp and port to modern OSes, is now AI slop. The author seems to have gotten AI psy…
↗Not briefed
hn
hn-ai
1moPairing Claude Code with a Local Agent on Surface Laptop Ultra—
↗Not briefed
bluesky
bluesky-ai
1moThe authors of this paper claims that they uploaded a draft paper to ChatGPT and OpenAI plagiarized their work arxiv.org/pdf/2610.10072🔧 ChatGPTThe authors of this paper claims that they uploaded a draft paper to ChatGPT and OpenAI plagiarized their work arxiv.org/pdf/2610.10072
↗Not briefed
hn
hn-ai
1moVosti: Specifying, Implementing, and Verifying Deterministic LLM Inference—
↗Not briefed
bluesky
bluesky-ai
1moand that's good and cool, but we're treating openai as laudable, giving them the credit for expanding human knowledge, but then it will be …and that's good and cool, but we're treating openai as laudable, giving them the credit for expanding human knowledge, but then it wil…
↗Not briefed
bluesky
bluesky-ai
1moI picked "relitigating" to emphasise that the LLM industry perpetually thinks computing began in 2018 and everything is new. I'm pretty sur…I picked "relitigating" to emphasise that the LLM industry perpetually thinks computing began in 2018 and everything is new. I'm prett…
↗Not briefed
arxiv
arxiv-ai
1moA Categorical Atlas of Parameter-Efficient Fine-Tuning: Cones, Orbits, and GaugesParameter-efficient fine-tuning has hundreds of methods, usually compared by parameter count and benchmark score. We compare them as m…
↗Not briefed
bluesky
bluesky-ai
1moApparently I need to repeat myself, because OpenAI clearly didn’t talk to mathematicians about their PR effort, presented it to the public …Apparently I need to repeat myself, because OpenAI clearly didn’t talk to mathematicians about their PR effort, presented it to the pu…
↗Not briefed
arxiv
arxiv-ai
1moZebra3D: Efficient Hybrid Foundation Models For Scalable 3D ReconstructionFeed-forward 3D foundation models reconstruct scenes from unposed images, but the quadratic cost of global attention limits their scal…
↗Not briefed
bluesky
bluesky-ai
1moI will never use Claude but I will insult and curse every chatbot/AI agent I am forced to deal with.I will never use Claude but I will insult and curse every chatbot/AI agent I am forced to deal with.
↗Not briefed
hn
hn-ai
1moMaiSus – An operating system built from scratch in 100% Assembly—
↗Not briefed
rss
simon-willison
1moQuoting Carson Gross—
↗Not briefed
hn
hn-ai
1moHuggingFace adds p(doom) option for user profiles—
↗Not briefed
arxiv
arxiv-ai
1moSpatial Context or Memory? Sizing the Inputs of Neural Surrogates for Partially Observed FieldsNeural surrogates that forecast a physical field from sparse, noisy sensors read a window of nearby sensors and a history of past snap…
↗Not briefed
bluesky
bluesky-ai
1moIs this goodIs this good
↗Not briefed
bluesky
bluesky-ai
1moIt’s not that I think Anthropic is right in treating an LLM better than the people who work for them or are victimized by their software, i…It’s not that I think Anthropic is right in treating an LLM better than the people who work for them or are victimized by their softwa…
↗Not briefed
hn
hn-ai
1moUsing GitHub Issues with a Screen Reader—
↗Not briefed
hn
hn-ai
1moShow HN: Free open source Adobe Lightroom alternative, completely local with AI—
↗Not briefed
rss
techcrunch-ai
1moPretend you’re sitting at Elizabeth Holmes’ desk on this weirdly detailed websiteWith over a thousand emails, slides, texts, and documents from the United States v. Elizabeth Holmes trial, Extend engineer Bo Lau cre…
↗Not briefed
bluesky
bluesky-ai
1moVesta gets $30M for mortgage agent swarms Vesta, an AI-native startup, raised $30M to deploy AI agents that automate mortgage origination …Vesta gets $30M for mortgage agent swarms Vesta, an AI-native startup, raised $30M to deploy AI agents that automate mortgage origina…
↗Not briefed
hn
hn-ai
1moShow HN: Brewlet is Java on Kubernetes with node-managed JDKs—
↗Not briefed
arxiv
arxiv-ai
1moA Distributional Framework for Behavioral Characterization and Risk Management of Large Language ModelsLarge language models are stochastic by nature. Repeated executions under the same experimental conditions may produce different respo…
↗Not briefed
bluesky
bluesky-ai
1moShocked Pikachu.Shocked Pikachu.
↗Not briefed