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
3dDecision Making Cannot Be Outsourced—
↗Not briefed
rss
techmeme
3dAn analysis of 350+ Amazon holiday deals: 29.1% were priced the same or higher than in the prior 30 days, and ~60% had no savings compared to the prior 90 days (Groundwork Collaborative)Groundwork Collaborative: An analysis of 350+ Amazon holiday deals: 29.1% were priced the same or higher than in the prior 30 days, an…
↗Not briefed
hn
hn-ai
3dActors in UK to get more rights over their 'digital replicas' in deal on AI—
↗Not briefed
bluesky
bluesky-ai
3d#andrewcampbell #britishartist #art #subvert #meta #agitate #algorithms #provoke #X #attack #openai #resist #ai #disturb #instagram #artfor…#andrewcampbell #britishartist #art #subvert #meta #agitate #algorithms #provoke #X #attack #openai #resist #ai #disturb #instagram #a…
↗Not briefed
arxiv
arxiv-ai
3dCOTTI: Contrastive Training and Time-Aware Inference for Robust Industrial Recommendation under Promotional DriftPromotional events shift traffic composition, purchase behavior, and the scale of conversion probabilities on different time horizons.…
↗Not briefed
bluesky
bluesky-ai
3dIt's worth paying attention not just to what AI can & can't do, but also how companies change the nature of human work in order to make AI …It's worth paying attention not just to what AI can & can't do, but also how companies change the nature of human work in order to mak…
↗Not briefed
arxiv
arxiv-ai
3dPropose with the Gradient, Accept with the Loss: One Hedged Attack for Pixel-Sparse RobustnessA pixel-sparse adversary may change a few pixels of an image to any value. Robustness to such attacks is currently measured with an en…
↗Not briefed
arxiv
arxiv-ai
3dLost in the Funnel: Selection-Induced Miscalibration in Ranking and Decision SystemsModern ranking and decision systems combine components with divergent objectives, and scores calibrated at the model boundary are used…
↗Not briefed
arxiv
arxiv-ai
3dPACT: Plan-Action Consistency Training by Preserving Correct Plans and Repairing Actions in GUI AgentsGUI agents can formulate a correct immediate plan in their reasoning yet emit an action that fails to execute it, creating an executio…
↗Not briefed
arxiv
arxiv-ai
3dAdaptive LLM-Driven Honeypots vs. LLM-Driven Attackers: Does Narrative Coherence Improve Deception?Large Language Models (LLMs) enable honeypots to generate responses and deceptive artefacts on demand, yet existing approaches evaluat…
↗Not briefed
rss
techmeme
3dJD.com reports Q3 revenue up 3.5% YoY to ~$46.4B, meeting est., and net income up 22% to ~$975M, amid intense competition and a sluggish Chinese economy (Rachel Yeo/Bloomberg)Rachel Yeo / Bloomberg: JD.com reports Q3 revenue up 3.5% YoY to ~$46.4B, meeting est., and net income up 22% to ~$975M, amid intense…
↗Not briefed
bluesky
bluesky-ai
3dMYCKET färre av problemen jag vill åtgärda finns för modeller man kör lokalt på sin dator som vilket program som helst. Där kan vi ha disku…MYCKET färre av problemen jag vill åtgärda finns för modeller man kör lokalt på sin dator som vilket program som helst. Där kan vi ha…
↗Not briefed
arxiv
arxiv-ai
3dEvolution Under Fire: Evolving Agent Safety Harnesses with Adaptive Injection StressorsTool-using language-model agents must act on information from untrusted emails, webpages, and documents. This creates a central securi…
↗Not briefed
hn
hn-ai
3dAsk HN: What's the new open source business modelYes that's a double negative because open source is not a business model. For years we've argued about value creation, susta…
↗Not briefed
bluesky
bluesky-ai
3dThe French CEO of Hugging Face, a mega-hub of open-source AI software, says the attack by OpenAI models earlier this year proved one thing:…🔧 Hugging FaceThe French CEO of Hugging Face, a mega-hub of open-source AI software, says the attack by OpenAI models earlier this year proved one t…
↗Not briefed
arxiv
arxiv-ai
3dAfriNORM: An Evaluation Benchmark for Normative Reasoning of LLMs Across African Cultures🔧 GPT-5Large language models (LLMs) are increasingly deployed across various sectors. However, current evaluation benchmarks are insufficient…
↗Not briefed
rss
bing-news-ai
3dUniversity of Phoenix leaders present on artificial intelligence, academic policy, leadership development and workplace innovation at PACRAO 2026University of Phoenix leaders presented on artificial intelligence, academic policy management, workplace flexibility and leadership d…
↗Not briefed
hn
hn-ai
3dA swimmable underwater world in one 97kb HTML file—
↗Not briefed
bluesky
bluesky-ai
3dReally enjoying trying a different drawing style for this #tattoo #commission I'm working on... #art #sciart #paleoart #mosasaurusReally enjoying trying a different drawing style for this #tattoo #commission I'm working on... #art #sciart #paleoart #mosasaurus
↗Not briefed
bluesky
bluesky-ai
3dあたらしい子お迎えするの…♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #小説 #AIイラスト #Yum…あたらしい子お迎えするの…♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #小説 #AIイラスト…
↗Not briefed
arxiv
arxiv-ai
3dReMaRL: Learning to Unmask and Remask in Frozen Diffusion Language ModelsMasked diffusion language models generate tokens in parallel, and the order in which tokens are unmasked affects final-answer accuracy…
↗Not briefed
arxiv
arxiv-ai
3dDo LLMs Understand Network Traces? A Comparative Study of Reasoning over Raw and Filtered Traffic DataThe effective delivery and monitoring of ICT services rely heavily on network traffic analysis. However, interpreting complex packet c…
↗Not briefed
rss
techmeme
3dChinese records: Alice Huang's husband Kang Wei built a state-backed data center network in China, with proposed investment exceeding $9B between 2021 and 2024 (Lingling Wei/Wall Street Journal)Lingling Wei / Wall Street Journal: Chinese records: Alice Huang's husband Kang Wei built a state-backed data center network in China,…
↗Not briefed
rss
wired-ai
3dRogue Agents Hacked His Company. Why Is Clément Delangue So Optimistic About AI?🔧 Hugging FaceThe French CEO of Hugging Face, a mega-hub of open-source AI software, says the attack by OpenAI models earlier this year proved one t…
↗Not briefed
rss
wired-ai
3dThe Apple Watch’s Live Rewind Can Record and Rewind Spoken Conversations. It’s Pretty AwkwardLive Rewind and Siri Recap can transcribe missed moments and summarize your day—if you don’t mind a little social clumsiness.
↗Not briefed
rss
stratechery
3dApple Permissions Update, The New App Paradigm, My iPhone Duo ReviewApple doesn't ruin the Mac, but they don't fix the agent problem, raising questions if they can think outside the app paradigm. Then,…
↗Not briefed
bluesky
bluesky-ai
3dThey accepted none of his bullshit. 5) He argued the plugin Claude built for him would make life so much easier. Designer pointed out it is…They accepted none of his bullshit. 5) He argued the plugin Claude built for him would make life so much easier. Designer pointed out…
↗Not briefed
hn
hn-ai
3dHow to Get Rich (2019)—
↗Not briefed
bluesky
bluesky-ai
3dReminder that Anthropic is a cult and their values are very likely not to align with yours, or with anyone’s.Reminder that Anthropic is a cult and their values are very likely not to align with yours, or with anyone’s.
↗Not briefed
arxiv
arxiv-ai
3dMillBill: A Machining Agent with Geometric FeedbackTurning a CAD model into a machining program requires both process knowledge and geometric precision. Language models can generate suc…
↗Not briefed
bluesky
bluesky-ai
3dFrom the very beginning, Anthropic has been that "will refuse and give you a moral lecture" company. Remember when it refused to help you k…From the very beginning, Anthropic has been that "will refuse and give you a moral lecture" company. Remember when it refused to help…
↗Not briefed
rss
techmeme
3dA look at Anthropic's data center network plans, dubbed Anthropic Direct, committing $900B+ through 2045; docs show it will lease 875K TPUs for $7.2B per year (Financial Times)Financial Times: A look at Anthropic's data center network plans, dubbed Anthropic Direct, committing $900B+ through 2045; docs show i…
↗Not briefed
arxiv
arxiv-ai
3dCADMAGEDDON: High-Fidelity CAD Reconstruction with Local InferenceCAD practitioners often work with proprietary parts that they are unwilling to share with external services. For individuals and small…
↗Not briefed
rss
wired-ai
3dAI Told These Ex-Google Guys How to Build a Billion-Dollar Weapons BusinessGoogle and Apple alumni Chad Steelberg and Andy Baynes are applying what they learned running software companies and designing iPhones…
↗Not briefed
hn
hn-ai
3dAncient Technology Makes These Water Bells Sing [video]—
↗Not briefed
hn
hn-ai
3dLooped Transformers Explained—
↗Not briefed
arxiv
arxiv-ai
3dBalancing Exploration and Exploitation in Constrained Combinatorial Optimization with U(1)-Symmetric Tensor TrainsTensor Train (TT) decomposition has emerged as a powerful tool for mitigating the curse of dimensionality in a variety of machine lear…
↗Not briefed
hn
hn-ai
3dShow HN: Wake, see what Claude Code did and where the tokens wentHi everyone, I'm Ricardo and I've made a (free & Open Source) macOS app for knowing where my Claude Code subscription to…
↗Not briefed
hn
hn-ai
3dShow HN: Mainfast – GitHub checks that catch merges that silently break main—
↗Not briefed
bluesky
bluesky-ai
3dCalifornia sues OpenAI for misleading public about AI safety www.theregister.com/ai-and-ml/20...California sues OpenAI for misleading public about AI safety www.theregister.com/ai-and-ml/20...
↗Not briefed
rss
techmeme
3dSources: Nvidia Developer Tech Director Nader Khalil wrote to board arguing that failing to reinstate Jensen Huang could cost Nvidia "irreplaceable talent" (Wall Street Journal)Wall Street Journal: Sources: Nvidia Developer Tech Director Nader Khalil wrote to board arguing that failing to reinstate Jensen Huan…
↗Not briefed
arxiv
arxiv-ai
3dPhysically Admissible Generative Flow Matching for Sparse Flow ReconstructionReconstructing full fluid velocity fields from sparse, localized observations is a severely ill-posed inverse problem: most of the spa…
↗Not briefed
bluesky
bluesky-ai
3dDet Silicon Valley lyckats bäst med är att likställa deras egna företag med tekniken själv. Som jämförelse bygger jag just nu en LLM som ka…Det Silicon Valley lyckats bäst med är att likställa deras egna företag med tekniken själv. Som jämförelse bygger jag just nu en LLM s…
↗Not briefed
hn
hn-ai
3dMPs ask what the break clause was for as Palantir wins two more years in NHS—
↗Not briefed
arxiv
arxiv-ai
3dTemporal Diffusion Models for Joint Signal-Channel Estimation under Gaussian and $α$-Stable MIMO Channels🔧 Stable DiffusionDiffusion models have emerged as powerful tools for learning complex data distributions and solving inverse problems. However, most di…
↗Not briefed
rss
marktechpost
3dNVIDIA Open-Sources NVKE and NVKR for Managed Kubernetes on GPU InfrastructureNVIDIA has open-sourced NVIDIA Kubernetes Engine (NVKE) and NVIDIA Kubernetes Recipes (NVKR) under Apache 2.0. Together they let AI fa…
↗Not briefed
arxiv
arxiv-ai
3dJointGV: Structure-Aware Correction of Visual World Models with Parallel Graph Dynamics for Long-Horizon ModelingVisual world models predict future visual states conditioned on actions, but recursive errors in entity poses and spatial relations li…
↗Not briefed
rss
techmeme
3dSources: Nvidia warns customers that the proposed Remote Access Security Act could subject chip buyers to US legal action if Chinese companies access AI chips (Kai Nicol-Schwarz/CNBC)Kai Nicol-Schwarz / CNBC: Sources: Nvidia warns customers that the proposed Remote Access Security Act could subject chip buyers to US…
↗Not briefed
arxiv
arxiv-ai
3dFrom Perception to Simulation: Unveiling and Empowering Mental Spatial Dynamics in MLLMsSpatial reasoning is a fundamental component of world understanding for Multimodal Large Language Models (MLLMs). While existing resea…
↗Not briefed
arxiv
arxiv-ai
3dJET: Just-Enough Exploration Tuning for Efficient Tool-Calling Agents in Repository-Level Question AnsweringIn repository-level question answering (QA), an agent explores a code repository with tools and should answer a question once enough i…
↗Not briefed