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
arxiv
arxiv-ai
12dHidden Tails: Certifying Tail-Risk Claims under Selective LabelsHistorical decisions determine which outcomes are observed, potentially concealing a predictor's largest errors. Even unlimited data f…
↗Not briefed
arxiv
arxiv-ai
12dA local language model layered with rule-based and neural detectors for clinical text de-identification: an open-source toolkit and evaluationObjectives: To measure whether a locally hosted large language model (LLM), stacked with deterministic and clinical named-entity recog…
↗Not briefed
bluesky
bluesky-ai
12dAll right gang, this might become a rant, we will have to see. My new manager has gone back to full AI is the future, embrace it or get lef…All right gang, this might become a rant, we will have to see. My new manager has gone back to full AI is the future, embrace it or ge…
↗Not briefed
bluesky
bluesky-ai
12dAnthropic joins AI’s natural gas frenzy, buys $3.4B worth of turbinesAnthropic joins AI’s natural gas frenzy, buys $3.4B worth of turbines
↗Not briefed
hn
hn-ai
12dShow HN: A visualization map of the funding of public companiesI made this webapp (vibecoded) with Meta's AI models during the Meta AI hackathon that just ended (Oct 26-Nov 4). So far it conta…
↗Not briefed
arxiv
arxiv-ai
12dMulti-Agent Screening: What Agreement and Acceptance Cannot Tell YouWhen AI agents screen items for a costly human assessor, as in agentic search, e-discovery and AI reviewing of papers, only what they…
↗Not briefed
arxiv
arxiv-ai
12dFisher-Glass Learning: Classification Bounds and Subpopulation AuditsAccurate classifiers can fail when background cues change without altering the correct labels. Predictive performance alone does not q…
↗Not briefed
hn
hn-ai
12dBeyond Artex: 45 Agentic Attack Frameworks, and How to Find Them—
↗Not briefed
arxiv
arxiv-ai
12dGeoScope: Training-Free Program Execution beyond the Counting Horizon for Ultra-High-Resolution Remote Sensing Question AnsweringIn ultra-high-resolution (UHR) remote sensing images, the objects that questions count in a region or relate by direction often span o…
↗Not briefed
rss
techcrunch-ai
12dAnthropic joins AI’s natural gas frenzy, buys $3.4B worth of turbinesAnthropic bought billions of dollars worth of natural gas turbines in a deal that could give it buyer's remorse.
↗Not briefed
arxiv
arxiv-ai
12dGeneralized Adaptive Boosting and the Geometry of MistakesNot all mistakes are alike. For example, in medical diagnosis, false positive and false negative mistakes have a different impact and…
↗Not briefed
arxiv
arxiv-ai
12dTACTUS: A Robotic Palpation System for Spatially Heterogeneous Hardness Mapping of Woody Breast MyopathyWoody breast (WB) myopathy is a spatially heterogeneous fibro-degenerative condition of the broiler Pectoralis major muscle. Existing…
↗Not briefed
arxiv
arxiv-ai
12dAgnostic Marginal-Nonuniform LearningWe study the problem of binary classification in the marginal-nonuniform learning framework in the agnostic setting. In this model, th…
↗Not briefed
rss
the-decoder
12dDisney bets on Adobe's Firefly to create brand-safe AI models for its entertainment empireDisney and Adobe are building custom AI tools together for the entertainment giant's productions. The article Disney bets on Adobe's F…
↗Not briefed
hn
hn-ai
12dShow HN: Divorce – an agent skill for native HTML, verified with Nu and axe—
↗Not briefed
bluesky
bluesky-ai
12dYesterday people were falling for ai slop saying that nobody at their company can read code anymore because of Claude and now today people …Yesterday people were falling for ai slop saying that nobody at their company can read code anymore because of Claude and now today pe…
↗Not briefed
bluesky
bluesky-ai
12ddo you think it'd be better or worse if he was having a relationship with an AI girlfriend built on deepseekdo you think it'd be better or worse if he was having a relationship with an AI girlfriend built on deepseek
↗Not briefed
arxiv
arxiv-ai
12dA Lattice-Based HMM Approach to Aruz Scansion of Ottoman PoetryOttoman Divan poetry is composed in aruz, a quantitative meter based on the weights of syllables. The meter is sensitive to specific c…
↗Not briefed
arxiv
arxiv-ai
12dAverage versus expected causal effects under interference: Modeling assumptions and estimation bottlenecksCausal inference under treatment interference requires assumptions to reduce the exponentially large number of potential outcomes. We…
↗Not briefed
arxiv
arxiv-ai
12dCoilStellaration: A Dataset and Benchmark for Engineering-Aware Stellarator Coilset GenerationStellarators are fusion devices that use externally generated, three-dimensional magnetic fields to confine hot plasma and offer a pro…
↗Not briefed
arxiv
arxiv-ai
12dreComposer: Editable Planning for Controllable Image GenerationPrompt-based image generation makes it easy to produce visual assets, but offers little support for engaging with the creative process…
↗Not briefed
hn
hn-ai
12dAn SEO page checker that shows its scoring and admits its limits—
↗Not briefed
bluesky
bluesky-ai
12dThe LLM vendors support export bans, because they want to be insulated from while they're at the topThe LLM vendors support export bans, because they want to be insulated from while they're at the top
↗Not briefed
arxiv
arxiv-ai
12dA four-step validation protocol for radiomic survival analysis in small cohorts: a body composition case studyGastroenteropancreatic neuroendocrine tumors (GEP-NET) lack prognostic biomarkers, and computed tomography (CT)-based body composition…
↗Not briefed
bluesky
bluesky-ai
12dBut if you're a Chinese LLM company, you still follow the path of least resistance: you want to catch up in LLM capability, not have your s…But if you're a Chinese LLM company, you still follow the path of least resistance: you want to catch up in LLM capability, not have y…
↗Not briefed
arxiv
arxiv-ai
12dLibrary of Hands: Fix the Hands, Not the World ModelA world model predicts how a scene changes after an action, and a planner uses it to search for actions that reach a goal. If the came…
↗Not briefed
arxiv
arxiv-ai
12dVLA Adaptation with Real and Simulated Data: A Case Study on Robotic ChessVision-language-action (VLA) models are trained as general-purpose foundation models for robot learning, and require fine-tuning for e…
↗Not briefed
arxiv
arxiv-ai
12dInterpretability Is Not Model-SpecificInterpretability tools are usually built for one model at a time: a refusal direction, a truthfulness probe, or a steering vector is f…
↗Not briefed
bluesky
bluesky-ai
12dCoPilot a Microsoft LLM still doesn't know the correct column names in the SQL Server SSIS Catalog, also a Microsoft technology that has ex…CoPilot a Microsoft LLM still doesn't know the correct column names in the SQL Server SSIS Catalog, also a Microsoft technology that h…
↗Not briefed
rss
latent-space
12dWhat Frontier Labs Buy: A Data Deep Dive with Handshake AI — Garrett Lord & TeamOur first deep dive with an AI data company. In roughly a year, Handshake built an AI training business doing nearly $1 billion in gro…
↗Not briefed
bluesky
bluesky-ai
12dSam Altman should know how we are wasting all that RAM.Sam Altman should know how we are wasting all that RAM.
↗Not briefed
hn
hn-ai
12dCICS Dungeon—
↗Not briefed
bluesky
bluesky-ai
12dCloudflare explains more about Act 4 revenue (which they revealed in early July) -> Cloudflare on how OpenAI pays for hints that tell its c…Cloudflare explains more about Act 4 revenue (which they revealed in early July) -> Cloudflare on how OpenAI pays for hints that tell…
↗Not briefed
hn
hn-ai
12dI Am a Silicon Valley Veteran. We Are Sabotaging Ourselves—
↗Not briefed
arxiv
arxiv-ai
12dControlled Binary Assignment FlowsThis paper introduces a geometric approach to the representation and learning of time-variant parametrizations of deep assignment flow…
↗Not briefed
bluesky
bluesky-ai
12dReminder that Trump's brain works much like an LLM and the things it makes him say in an interview do not convey an actual belief, just a c…Reminder that Trump's brain works much like an LLM and the things it makes him say in an interview do not convey an actual belief, jus…
↗Not briefed
hn
hn-ai
12dDBDelve: Open-Source DB Client Made with Rust and GPUIHi HN, I’m Shayan and a few months ago I started building a DB Client DBDelve with a focus on performance and just making something th…
↗Not briefed
arxiv
arxiv-ai
12dTrade Promotion Optimisation for Consumer GoodsTrade promotions consume a large share of consumer-goods revenue, yet most planners still set discounts with a single, often unreliabl…
↗Not briefed
bluesky
bluesky-ai
12dAnthropic must face lawsuit over Claude subscription marketing reut.rs/4Ag2EpSAnthropic must face lawsuit over Claude subscription marketing reut.rs/4Ag2EpS
↗Not briefed
rss
marktechpost
12dBaidu’s ERNIE Evolve Automated Research System Beats NanoChat and NanoGPT RecordsBaidu’s ERNIE Evolve is an automated research system that edits code, runs experiments and keeps only verified gains. It beat public b…
↗Not briefed
bluesky
bluesky-ai
12ddo #gamedev people see using AI as all encompassing, like, you use claude to help code, or you use it for graphics, sounds, models, and art…do #gamedev people see using AI as all encompassing, like, you use claude to help code, or you use it for graphics, sounds, models, an…
↗Not briefed
arxiv
arxiv-ai
12dCertify, Report, or Refuse: Harm-Constrained Conformal Prediction for Clinical ActionsRouting a melanoma case to routine follow-up is a serious action error that diagnostic coverage alone does not rule out. We present Ha…
↗Not briefed
hn
hn-ai
12dMalloyBench: Language models write Malloy well with its compiler—
↗Not briefed
bluesky
bluesky-ai
12d@cdelawalla.bsky.social Especially important for HHS/RFK Jr since several public reports (transgender, nutrition) were published w/evidence…@cdelawalla.bsky.social Especially important for HHS/RFK Jr since several public reports (transgender, nutrition) were published w/evi…
↗Not briefed
hn
hn-ai
12dClaude – just do my resume—
↗Not briefed
bluesky
bluesky-ai
12dLLMに株価分析させるやつ、取引が活発でなくて単位元の額が高いもの(そんなものはあるのか?)であれば個人でも勝てるんじゃないかという気がしてきたな、流動性が低いものは機関投資家が取引しないだろうから割安のまま放置されてるだろうし、同じようなことをしてる個人投資家はそもそも買える…LLMに株価分析させるやつ、取引が活発でなくて単位元の額が高いもの(そんなものはあるのか?)であれば個人でも勝てるんじゃないかという気がしてきたな、流動性が低いものは機関投資家が取引しないだろうから割安のまま放置されてるだろうし、同じようなことをしてる個人投資家はそも…
↗Not briefed
bluesky
bluesky-ai
12dIf you ever have to explain what AI psychosis is to someone just tell them it's when an LLM plays a game of Keepie-uppy with your Dunning K…If you ever have to explain what AI psychosis is to someone just tell them it's when an LLM plays a game of Keepie-uppy with your Dunn…
↗Not briefed
hn
hn-ai
12dData intelligence in your app is as easy as a Git checkout—
↗Not briefed
bluesky
bluesky-ai
12dAnd not least because those logs are almost certainly already in the hands of foreign governments! LLM use is the greatest gift to spycraft…And not least because those logs are almost certainly already in the hands of foreign governments! LLM use is the greatest gift to spy…
↗Not briefed
rss
techmeme
12dSources: a Commerce Department investigation into Megaspeed for allegedly smuggling Nvidia chips to China was transferred to the DOJ after Nvidia intervened (New York Times)New York Times: Sources: a Commerce Department investigation into Megaspeed for allegedly smuggling Nvidia chips to China was transfer…
↗Not briefed