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
1mo知人と「そのうちパッケージアプリって簡単にクローンできるようになるかもね」と話してたら、Adobe CCのスケールで始まってた。 逆コンパイル/逆アセンブルみたいな強引な手法に頼らなくても、提供される機能とUIとマニュアルが手に入れば、LLMで商用アプリケーションのクローンが…知人と「そのうちパッケージアプリって簡単にクローンできるようになるかもね」と話してたら、Adobe CCのスケールで始まってた。 逆コンパイル/逆アセンブルみたいな強引な手法に頼らなくても、提供される機能とUIとマニュアルが手に入れば、LLMで商用アプリケーションの…
↗Not briefed
arxiv
arxiv-ai
1moTurboPairFormer: Fast and Stable Protein Folding Model Training with an Optimized Triangle Attention KernelTriangular attention is a core computation in AlphaFold3-style biomolecular models, with cubic cost in token count. Its shared pair bi…
↗Not briefed
arxiv
arxiv-ai
1moThe Blind Spot Paradox: When Adaptive Classifiers Defeat Drift DetectorsMonitoring concept drift from an adaptive classifier's error stream creates an operational conflict with the model's own update loop.…
↗Not briefed
arxiv
arxiv-ai
1moRepICL: Reusable In-Context Prediction Across Heterogeneous Representation SpacesFrozen representations are widely reused for downstream classification, yet each new task typically requires fitting a new predictor.…
↗Not briefed
bluesky
bluesky-ai
1mocounterpoint: google's default search LLM cannot distinguish a batch vs sequential unit operation when parsed from a simple 2 sentence summ…counterpoint: google's default search LLM cannot distinguish a batch vs sequential unit operation when parsed from a simple 2 sentence…
↗Not briefed
rss
techmeme
1moSam Altman says he's "very uncomfortable" with attributing religious force to AI, calling it "a real safety issue", as Anthropic engages with religious leaders (Ben Berkowitz/Axios)Ben Berkowitz / Axios: Sam Altman says he's “very uncomfortable” with attributing religious force to AI, calling it “a real safety iss…
↗Not briefed
bluesky
bluesky-ai
1morozhl.as/lLM Šéfredaktorky a šéfredaktoři čtyřiceti českých médií společně podepsali dokument adresovaný politikům. „Vyzýváme proto veřejné…rozhl.as/lLM Šéfredaktorky a šéfredaktoři čtyřiceti českých médií společně podepsali dokument adresovaný politikům. „Vyzýváme proto ve…
↗Not briefed
rss
techmeme
1moOKXICE, a JV between OKX and ICE, is seeking SEC approval to offer tokenized shares of 63 NYSE-listed firms, as it leverages new US rules on tokenized stocks (Scott Patterson/Bloomberg)Scott Patterson / Bloomberg: OKXICE, a JV between OKX and ICE, is seeking SEC approval to offer tokenized shares of 63 NYSE-listed fir…
↗Not briefed
bluesky
bluesky-ai
1moThe meat proxy tactic online is wild. Quickly replacing the "I googled it and..." nonsense. We all have access to Google and [insert whate…The meat proxy tactic online is wild. Quickly replacing the "I googled it and..." nonsense. We all have access to Google and [insert…
↗Not briefed
arxiv
arxiv-ai
1moCACFG: Curvature-Aware Classifier-Free Guidance and Optimal ControlDiffusion models generate samples by learning to reverse a fixed corruption process, and classifier-free guidance (CFG) is the standar…
↗Not briefed
bluesky
bluesky-ai
1mo多少なりとも期待した自分がバカでしたね。 少なくともOpenAIに関しては何も変わっていない。多少なりとも期待した自分がバカでしたね。 少なくともOpenAIに関しては何も変わっていない。
↗Not briefed
bluesky
bluesky-ai
1mo名作絵本『がまくんとかえるくん』風にAIの暴走によるハッキングを風刺した『がまくんとかえるくんとますます進化する機械たち』 https://gigazine.net/news/20261005-frogandtoad-ai-openai-hugging-face-hack/名作絵本『がまくんとかえるくん』風にAIの暴走によるハッキングを風刺した『がまくんとかえるくんとますます進化する機械たち』 https://gigazine.net/news/20261005-frogandtoad-ai-openai-hugging-face-ha…
↗Not briefed
bluesky
bluesky-ai
1moPretty hard to argue that critics and AI labs are simply “talking past each other” when one side is clearly insane. www.politico.com/news/2…Pretty hard to argue that critics and AI labs are simply “talking past each other” when one side is clearly insane. www.politico.com/n…
↗Not briefed
bluesky
bluesky-ai
1mo#Trump has dismissed calls for stronger oversight and largely left it to #AI companies to regulate themselves. The pact signed on 29 Septem…#Trump has dismissed calls for stronger oversight and largely left it to #AI companies to regulate themselves. The pact signed on 29 S…
↗Not briefed
arxiv
arxiv-ai
1moPhaseMatcher: Autoregressive Phase-Set Identification with Spectral DecompositionRecovering complete phase sets from powder X-ray diffraction (PXRD) is challenging when weak-phase peaks overlap stronger signals. A n…
↗Not briefed
arxiv
arxiv-ai
1moAnchorPose for Geometry-Aware MOF Assembly through Meso-Grained Pose GenerationPredicting metal-organic framework (MOF) structures from given building blocks requires recovering their positions and orientations in…
↗Not briefed
bluesky
bluesky-ai
1moAlready announced it on Instagram and had to bring it to Blue sky as well: My upcoming LLM AIbert Einstein 10.0 will wipe them all off the…Already announced it on Instagram and had to bring it to Blue sky as well: My upcoming LLM AIbert Einstein 10.0 will wipe them all of…
↗Not briefed
arxiv
arxiv-ai
1moHLA: Expressive Hybrid Linear Attention via Chunk-Wise Dynamic MixingLinear attention enables efficient long-context autoregressive decoding by compressing history into recurrent states, but this compres…
↗Not briefed
bluesky
bluesky-ai
1moLLMでリバースエンジニアリングしたのかな。すっげー。マニュアルのPDFを読み込ませて、「このアプリを作れ」とか。 coliss.com/wp-content/c...LLMでリバースエンジニアリングしたのかな。すっげー。マニュアルのPDFを読み込ませて、「このアプリを作れ」とか。 coliss.com/wp-content/c...
↗Not briefed
arxiv
arxiv-ai
1moAdaptive-Shot Hybrid Quantum Anomaly Detection for Tactile Internet Security: Reliability-Aware Measurement Allocation Under Resource ConstraintsTactile Internet (TI) security analytics must balance reliable thresholded decisions with constrained computational and measurement re…
↗Not briefed
arxiv
arxiv-ai
1moHiER-BLS: A Hierarchy-Guided and Error-Correcting Robust Incremental Broad Learning SystemBroad Learning System (BLS) supports analytical training and incremental expansion, but its growth needs guidance on which inputs new…
↗Not briefed
arxiv
arxiv-ai
1moRequest Order Matters: Cache-History Sensitivity in Selective KV-Cache Reuse for Rolling AgentsLong-running agents repeatedly call an LLM while retaining most of their document window, evicting old documents, and appending new on…
↗Not briefed
bluesky
bluesky-ai
1moAIの社会実装の障害となっているのは、開発企業のこういったクソふざけた態度なのでは。 【OpenAIのサム・アルトマンCEOが「AIの恩恵に預かるために多少の悪いことは受けれるべき」との見解を示す 】 gigazine.net/news/2026100...AIの社会実装の障害となっているのは、開発企業のこういったクソふざけた態度なのでは。 【OpenAIのサム・アルトマンCEOが「AIの恩恵に預かるために多少の悪いことは受けれるべき」との見解を示す 】 gigazine.net/news/2026100...
↗Not briefed
rss
bing-news-ai
1moWhat’s the difference between artificial intelligence and superintelligence?The distinction between artificial intelligence and superintelligence matters because the capabilities of today’s AI systems are not t…
↗Not briefed
rss
bing-news-ai
1moWhat’s the difference between artificial intelligence and superintelligence?The distinction between artificial intelligence and superintelligence matters because the capabilities of today’s AI systems are not t…
↗Not briefed
rss
bing-news-ai
1moWhat’s the difference between artificial intelligence and superintelligence?The distinction between artificial intelligence and superintelligence matters because the capabilities of today’s AI systems are not t…
↗Not briefed
rss
bing-news-ai
1moWhat’s the difference between artificial intelligence and superintelligence?The distinction between artificial intelligence and superintelligence matters because the capabilities of today’s AI systems are not t…
↗Not briefed
arxiv
arxiv-ai
1moSelecting Long-Horizon Trajectories for Reliable and Efficient Terminal-Agent TrainingTerminal agents are commonly trained by imitating long teacher trajectories, yet how much of each trajectory to supervise remains unex…
↗Not briefed
arxiv
arxiv-ai
1moMeasurement-First Auditing of Agentic Leaderboards: Contamination Susceptibility, Matched-Control Re-evaluation, and Scorer Validation🔧 ConsensusAgentic leaderboards increasingly evaluate systems on public benchmarks whose task statements and solution-bearing artifacts can remai…
↗Not briefed
hn
hn-ai
1moNext-Generation Collective Communication Object for AMD GPUs—
↗Not briefed
bluesky
bluesky-ai
1moOpenAIのサム・アルトマンCEOが「AIの恩恵に預かるために多少の悪いことは受けれるべき」との見解を示す https://gigazine.net/news/20261005-sam-altman-talk-to-decoded/OpenAIのサム・アルトマンCEOが「AIの恩恵に預かるために多少の悪いことは受けれるべき」との見解を示す https://gigazine.net/news/20261005-sam-altman-talk-to-decoded/
↗Not briefed
arxiv
arxiv-ai
1moData-Driven Personas for Survey Simulation: Insights into Simulation Alignment Across Data-Access RegimesHowever, many existing steering approaches rely on target-domain human data for fine-tuning or prompting that is costly to collect and…
↗Not briefed
arxiv
arxiv-ai
1moUsefulness of Quantile-Aware Diffusion Modeling for Highly Imbalanced Tabular DataClassification problem in the context of highly imbalanced data is a major challenge in many real-world applications (e.g., FinTech, h…
↗Not briefed
hn
hn-ai
1moWeb games created with Claude Opus—
↗Not briefed
bluesky
bluesky-ai
1moMaybe ask an LLM?Maybe ask an LLM?
↗Not briefed
bluesky
bluesky-ai
1moI think consciousness is a spectrum. That said, a gnat is probably more conscious than any extant agentic LLM.I think consciousness is a spectrum. That said, a gnat is probably more conscious than any extant agentic LLM.
↗Not briefed
hn
hn-ai
1moShow HN: A tool which estimates X statisticsI've built a tool that uses the official X algo with data scraped from X that estimates X engagement and helps you optimise your…
↗Not briefed
arxiv
arxiv-ai
1moWho Keeps the Gains from Personal AI Assistants? Seller Adaptation and the Unassisted in a Language-Model Market SimulationPersonal AI assistants are beginning to transact for consumers, and early adopters capture real savings. Whether those savings survive…
↗Not briefed
bluesky
bluesky-ai
1moLiehittelevät tietysti juuri paavia koska hänellä 1) on keskitettyä valtaa 2) protestantit painottaisivat, että tekoälyn on itsenäisesti tu…Liehittelevät tietysti juuri paavia koska hänellä 1) on keskitettyä valtaa 2) protestantit painottaisivat, että tekoälyn on itsenäises…
↗Not briefed
bluesky
bluesky-ai
1mooops we forgot to give it any benefitsoops we forgot to give it any benefits
↗Not briefed
arxiv
arxiv-ai
1moTransporting Unsecured Stacked Payloads with a Quadrupedal Robot via Multi-Objective Reinforcement LearningTransporting unsecured payloads with legged robots over uneven terrain requires balancing locomotion performance and payload stability…
↗Not briefed
arxiv
arxiv-ai
1moNash Equilibrium Text: A Game-Theoretic Decoding Framework for Text GenerationText revision has become an integral component of large language models. This paper formulates revision such that it admits a Nash equ…
↗Not briefed
arxiv
arxiv-ai
1moLevel-of-Token DiffusionImage and video diffusion models allocate equal computation to every region, even when the intended scene calls for varying levels of…
↗Not briefed
arxiv
arxiv-ai
1moPlan Canvas: Fixed Reasoning Regions for Continuous Language FlowsContinuous language flows generate text by denoising all positions of a target canvas together. The natural way to add reasoning to su…
↗Not briefed
bluesky
bluesky-ai
1moWhy a decade of AI safety promises keep failing to bite #ausnews #vicnews #ai #artificialintelligence #si #trump #maga #gop #seppos #altman…🔧 ChatGPTWhy a decade of AI safety promises keep failing to bite #ausnews #vicnews #ai #artificialintelligence #si #trump #maga #gop #seppos #a…
↗Not briefed
arxiv
arxiv-ai
1moFeature identification for parameter extraction and defect detection using machine learningProcess control of advanced semiconductor nodes is not only pushing the limits of metrology equipment requirements in terms of resolut…
↗Not briefed
rss
bing-news-ai
1moElon Musk says SpaceX AI could get a new name as Trump pushes for ‘Super Intelligence’SpaceX AI could soon have a different name as Musk responds to Trump’s push for the ‘Super Intelligence’ branding.
↗Not briefed
rss
bing-news-ai
1moElon Musk says SpaceX AI could get a new name as Trump pushes for ‘Super Intelligence’SpaceX AI could soon have a different name as Musk responds to Trump’s push for the ‘Super Intelligence’ branding.
↗Not briefed
rss
bing-news-ai
1moElon Musk says SpaceX AI could get a new name as Trump pushes for ‘Super Intelligence’SpaceX AI could soon have a different name as Musk responds to Trump’s push for the ‘Super Intelligence’ branding.
↗Not briefed
rss
bing-news-ai
1moElon Musk says SpaceX AI could get a new name as Trump pushes for ‘Super Intelligence’SpaceX AI could soon have a different name as Musk responds to Trump’s push for the ‘Super Intelligence’ branding.
↗Not briefed