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
2moYep, this Hacker News post has the right of it. They're pushing for a "great slowdown" because the gains have been incremental for, uh, a w…Yep, this Hacker News post has the right of it. They're pushing for a "great slowdown" because the gains have been incremental for, uh…
↗Not briefed
bluesky
bluesky-ai
2moYandex released the weights of a model (AliceAI-T5-35B-A0.6B-Base) used to generate AI answers in Search. Blog: medium.com/yandex/meet-...…Yandex released the weights of a model (AliceAI-T5-35B-A0.6B-Base) used to generate AI answers in Search. Blog: medium.com/yandex/mee…
↗Not briefed
hn
hn-ai
2moGPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?🔧 GPT-5—
↗Not briefed
bluesky
bluesky-ai
2moMore: www.ms.now/news/trump-a...More: www.ms.now/news/trump-a...
↗Not briefed
arxiv
arxiv-ai
2moClosing the Loop: Branch-and-Bound for Scalable Verification of Nonlinear Neural Feedback SystemsDespite recent advances in the verification of nonlinear neural feedback systems, scalability remains the central obstacle, as state-o…
↗Not briefed
bluesky
bluesky-ai
2moSeptember 10: Trump privately meets with OpenAI CEO Sam Altman September 14: Trump posts multiple times about how AI should not be regulat…September 10: Trump privately meets with OpenAI CEO Sam Altman September 14: Trump posts multiple times about how AI should not be re…
↗Not briefed
hn
hn-ai
2moMaking a vintage LLM from scratch; Take #2—
↗Not briefed
arxiv
arxiv-ai
2moIntelligent Interaction Techniques (IIxT) - ProposalInteraction techniques (IxTs) are the low-level, reusable components out of which user interfaces are designed, including menus, scrol…
↗Not briefed
hn
hn-ai
2moShow HN: Local historical wind and gust visualizerQuite a niche project that shows in a neat way historical wind force and orientation, anywhere on earth. No particular use case in min…
↗Not briefed
hn
hn-ai
2moShow HN: Pull every comment out of a Google Sheet, in the browserI spend a lot of time working in Google sheets. Often, I will collaborate with up to 50 other people in the same workbook. One specifi…
↗Not briefed
arxiv
arxiv-ai
2moNationally Consistent, Locally Incomplete: A Bayesian Remote-Sensing Audit of Rooftop Photovoltaic RegistriesTracking the energy transition requires reliable statistics on renewable deployment. Rooftop photovoltaics (PV) are especially hard to…
↗Not briefed
hn
hn-ai
2moShow HN: Build2me – Prove2Me for software agent swarms build codeHi everyone, this is my first original post on Hacker News. I’d like to share my thoughts on the next generation of AI agents. I belie…
↗Not briefed
bluesky
bluesky-ai
2mothis pattern already plays out in microcosm in the llm release cycle tbh. a new one comes out that saturates more benchmarks & doesn't go o…this pattern already plays out in microcosm in the llm release cycle tbh. a new one comes out that saturates more benchmarks & doesn't…
↗Not briefed
hn
hn-ai
2moAgentic Coding Strains CI: Scaling Test Impact Analysis at Anthropic—
↗Not briefed
bluesky
bluesky-ai
2moTrump privately met with OpenAI’s Altman at GOP convention, MS Now reports reut.rs/46WZKcZTrump privately met with OpenAI’s Altman at GOP convention, MS Now reports reut.rs/46WZKcZ
↗Not briefed
bluesky
bluesky-ai
2moAnthropic CEO & Co-Founder Dario Amodei called for a global slowdown of AI development to focus on safety through independent evaluators, g…Anthropic CEO & Co-Founder Dario Amodei called for a global slowdown of AI development to focus on safety through independent evaluato…
↗Not briefed
bluesky
bluesky-ai
2mowww.newsweek.com/what-chuck-s...www.newsweek.com/what-chuck-s...
↗Not briefed
arxiv
arxiv-ai
2moSymmetric solution of the Bellman optimality equation for repeated harmony gameIn social dilemma games, additional rewards or punishments have been studied as means of promoting cooperation. Therefore, it is impor…
↗Not briefed
arxiv
arxiv-ai
2moDrift Field Net: Learning Ocean Lagrangian advection fields from in-situ and satellite observationsThe North Pacific Subtropical Gyre (NPSG) is a major accumulation zone for floating plastic debris, resulting from basin-scale converg…
↗Not briefed
bluesky
bluesky-ai
2moIn this story Anthropic told partners that "gross margins are above 80% before accounting for revenue shared with distribution partners, in…In this story Anthropic told partners that "gross margins are above 80% before accounting for revenue shared with distribution partner…
↗Not briefed
hn
hn-ai
2moSteps for Writing a Speed-of-Light GEMM Kernel—
↗Not briefed
bluesky
bluesky-ai
2mothis was half of my interactions at VCFMW this weekend btw multiple times talking to someone about reverse engineering hardware or rewriti…this was half of my interactions at VCFMW this weekend btw multiple times talking to someone about reverse engineering hardware or re…
↗Not briefed
arxiv
arxiv-ai
2moProtoLIP: From Sentence-Level to Object-Level Evidence DisentanglementQuery-conditioned vision--language models enable fine-grained interpretation by revealing how visual evidence changes with textual que…
↗Not briefed
arxiv
arxiv-ai
2moDifferentially Private Semantic Plans for Aggregate Insight Generation\texttt{URANIA} provides end-to-end differential privacy (DP) for summaries of data-dependent clusters. However, its cluster--keyword…
↗Not briefed
arxiv
arxiv-ai
2moScaling Laws for Physics-Aware ACOPF Surrogate LearningLearning-based surrogates for AC optimal power flow (ACOPF) promise large speedups over classical solvers, but their operational value…
↗Not briefed
arxiv
arxiv-ai
2moSemantic-Aware Neural Video Codec for Error-Resilient Low-Latency TransmissionEmerging physical AI systems require low-latency, task-oriented video communication over unreliable channels. We propose a semantic-aw…
↗Not briefed
bluesky
bluesky-ai
2moClaude Fable solves 370-year cipher 🤖 Claude Fable 5.1 solved Sir Thomas Urquhart’s 370-year-old Cyphral Distich, a cryptogram that had e…Claude Fable solves 370-year cipher 🤖 Claude Fable 5.1 solved Sir Thomas Urquhart’s 370-year-old Cyphral Distich, a cryptogram that…
↗Not briefed
bluesky
bluesky-ai
2moWhat Amodei, Altman and Musk have said about AI risks, stoking 'doom' fears reut.rs/4rflJVZWhat Amodei, Altman and Musk have said about AI risks, stoking 'doom' fears reut.rs/4rflJVZ
↗Not briefed
arxiv
arxiv-ai
2moSpurious Tool Use: When RL Agents Learn the Wrong Reason to ActLarge language model (LLM) agents increasingly interleave natural language reasoning with external tools such as web search and code e…
↗Not briefed
bluesky
bluesky-ai
2moWitchy friends: what if… a rune or something to put in published works that, if the work is fed to an LLM, curses those responsible a littl…Witchy friends: what if… a rune or something to put in published works that, if the work is fed to an LLM, curses those responsible a…
↗Not briefed
hn
hn-ai
2moImageArm: Native macOS app to batch optimize images—
↗Not briefed
hn
hn-ai
2moShow HN: MCUHex – Flash and Live Monitoring for STM32 Devices over SWDI built MCUHex for monitoring global&#x2F;static variables and flashing Arm Cortex-M devices over SWD protocol.<p>Connect ST-Link deb…
↗Not briefed
bluesky
bluesky-ai
2moWarn people about AI risks, explain what could go wrong, and make the strongest argument you can, but please stop making up numbers.Warn people about AI risks, explain what could go wrong, and make the strongest argument you can, but please stop making up numbers.
↗Not briefed
bluesky
bluesky-ai
2moThis will allow us to hold an AI system hostage with a gun to its head, in the hope that all the other AI systems care enough about their b…This will allow us to hold an AI system hostage with a gun to its head, in the hope that all the other AI systems care enough about th…
↗Not briefed
arxiv
arxiv-ai
2moThe record is part of the task: matched-record evaluation of text classifiers across maintenance, safety and recall reportingMany operational cases are documented more than once, at different workflow stages and for different purposes, yet model evaluations n…
↗Not briefed
bluesky
bluesky-ai
2moAlso, like, individual personal duplicity is hardly necessary, not publicly commenting on employer activities is very normal. Frankly, Ant…Also, like, individual personal duplicity is hardly necessary, not publicly commenting on employer activities is very normal. Frankly…
↗Not briefed
arxiv
arxiv-ai
2moTowards Surrogate Based Dequantization of Quantum Reinforcement LearningIn recent years, the utility of parameterized quantum circuits as function approximators has been widely studied. In the context of re…
↗Not briefed
bluesky
bluesky-ai
2mo"As several people immediately asked: how many conjectures went into the hopper to get ten out? Nobody outside the lab knows, and that undi…"As several people immediately asked: how many conjectures went into the hopper to get ten out? Nobody outside the lab knows, and that…
↗Not briefed
rss
techmeme
2moAnthropic debuts Claude for Financial Advisors, with connectors to investment analytics and wealth-management tools from BlackRock, Addepar, Schwab, and others (Harshita Mary Varghese/Reuters)Harshita Mary Varghese / Reuters: Anthropic debuts Claude for Financial Advisors, with connectors to investment analytics and wealth-m…
↗Not briefed
bluesky
bluesky-ai
2mousing an internal LLM in my brain, if you willusing an internal LLM in my brain, if you will
↗Not briefed
arxiv
arxiv-ai
2moCompute-Optimal Pretrain--Fine-tune in Ridge Gradient DescentPretraining followed by fine-tuning introduces a compute-allocation problem: under a fixed training budget, compute spent improving th…
↗Not briefed
arxiv
arxiv-ai
2moMapping U.S. Federal AI Governance Against Sector VulnerabilityArtificial intelligence (AI) poses different levels of risk across sectors, but are these differences reflected in U.S. federal AI gov…
↗Not briefed
arxiv
arxiv-ai
2moThe AI-Enabled Scientific FrontierAs artificial intelligence's capabilities improve, it is increasingly viewed as a general scientific method. But how true are these cl…
↗Not briefed
arxiv
arxiv-ai
2moEfficient Reasoning Distillation: Small Video-Language Models via Synthetic CoT and Difficulty-Aware Fine-TuningWe present an efficient method to distill reasoning capabilities into compact video-language models (VLMs) for video question answerin…
↗Not briefed
arxiv
arxiv-ai
2moCADWorld: Computer-Use Benchmark for Long-Horizon Computer-Aided DesignComputer-use agents are increasingly evaluated in realistic desktop environments, but existing benchmarks provide limited coverage of…
↗Not briefed
bluesky
bluesky-ai
2monightmare blunt rotation at this "pro human" AI event tomorrow. Frank McCourt is a billionaire who has funded a ton of the dangerous and m…nightmare blunt rotation at this "pro human" AI event tomorrow. Frank McCourt is a billionaire who has funded a ton of the dangerous…
↗Not briefed
bluesky
bluesky-ai
2moright but the thing about piloting an LLM swarm to write agent<->agent interfaces/prompting is you still gotta _pilot_ itright but the thing about piloting an LLM swarm to write agent<->agent interfaces/prompting is you still gotta _pilot_ it
↗Not briefed
bluesky
bluesky-ai
2moDer US-Präsident weist Forderungen nach einer KI-Pause schroff zurück. Das nutze nur China. Die Debatte überlagert den geplanten Trump-Xi-G…Der US-Präsident weist Forderungen nach einer KI-Pause schroff zurück. Das nutze nur China. Die Debatte überlagert den geplanten Trump…
↗Not briefed
bluesky
bluesky-ai
2moAnother AI horror story dropped. Anthropic found 5 instances where government-linked accounts tried to use Claude for work on developing b…Another AI horror story dropped. Anthropic found 5 instances where government-linked accounts tried to use Claude for work on develop…
↗Not briefed
bluesky
bluesky-ai
2moIn the meantime, companies say they are going to start taking voluntary steps to reestablish public trust that they are acting responsibly.…In the meantime, companies say they are going to start taking voluntary steps to reestablish public trust that they are acting respons…
↗Not briefed