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
17dEvolutionary Adaptive Neighborhood Learning: A Genetic Algorithm for Instance-Dependent k Optimization in Nearest Neighbor ClassificationThe number of neighbours k is the central hyperparameter of the k-nearest neighbour (k-NN) classifier and is usually selected globally…
↗Not briefed
hn
hn-ai
17dOne TypeScript codebase, compiled native–from microcontrollers to consoles—
↗Not briefed
hn
hn-ai
17dAsk HN: Have you changed your setup in the age of AI and mass vulnerabilities?I'm curious if given all the recent advancements in LLMs but also news about hacks, vulnerabilities and exploits anyone has chang…
↗Not briefed
rss
bing-news-ai
17dMedicalAid.ai: How Artificial Intelligence Is Revolutionising Medical Aid Comparisons in South AfricaMedicalAid.ai introduces artificial intelligence to help South Africans compare complex medical aid schemes, matching plans to househo…
↗Not briefed
rss
bing-news-ai
17dMedicalAid.ai: How Artificial Intelligence Is Revolutionising Medical Aid Comparisons in South AfricaMedicalAid.ai introduces artificial intelligence to help South Africans compare complex medical aid schemes, matching plans to househo…
↗Not briefed
hn
hn-ai
17dZgui: Native Rust UI with fine-grained signals, CSS, and a retained GPU renderer—
↗Not briefed
bluesky
bluesky-ai
17dI think Anthropic should open the taps on giant virus research if they're looking for things that are horribly underfunded, hard for humans…I think Anthropic should open the taps on giant virus research if they're looking for things that are horribly underfunded, hard for h…
↗Not briefed
arxiv
arxiv-ai
17dUsing Transformers for Teaching and Learning via Symbolic Distillation of Neural A-life AgentsNeural networks controlling behavior of A-life agents can evolve over multiple generations. While effective in making the agents survi…
↗Not briefed
bluesky
bluesky-ai
17dVan Hollen as Senate leader would be a good pick. Also plenty of others. I would never trust any commitment from Schumer, we know where his…Van Hollen as Senate leader would be a good pick. Also plenty of others. I would never trust any commitment from Schumer, we know wher…
↗Not briefed
bluesky
bluesky-ai
17dThanks for your input, LLM chatbotThanks for your input, LLM chatbot
↗Not briefed
bluesky
bluesky-ai
17dおはよう世界👼🏼🪽 うぅ…ねむい…まだ朝…😮‍💨 【 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 #…おはよう世界👼🏼🪽 うぅ…ねむい…まだ朝…😮‍💨 【 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…
↗Not briefed
rss
huggingface-blog
17dA Closer Look at Phonon TTS. Launching High-Quality On-Device Speech That Tops the Benchmarks.—
↗Not briefed
bluesky
bluesky-ai
17dSome non-trivial amount of my resistance to AI is kind of a rebuke to the insult that I should pay a price for using a model, trained on th…Some non-trivial amount of my resistance to AI is kind of a rebuke to the insult that I should pay a price for using a model, trained…
↗Not briefed
bluesky
bluesky-ai
17dSorry, AI Bros, you don't get to go back now. Shoulda slipped Elon some of the good stuff and let him shuffle off but instead you lashed yo…Sorry, AI Bros, you don't get to go back now. Shoulda slipped Elon some of the good stuff and let him shuffle off but instead you lash…
↗Not briefed
arxiv
arxiv-ai
17dGoal Revelation as a Bridge Between Information-Theoretic Empowerment and Capability-Based Human PowerPower matters for AI safety. Optimal agents tend to acquire power across a wide range of objectives (instrumental convergence). As we…
↗Not briefed
arxiv
arxiv-ai
17dKG-Neighbor: Knowledge-Guided Neighborhood Construction for Embedding VisualizationIn embedding visualizations, nearby instances are interpreted as semantically related, although neighborhoods derived from input embed…
↗Not briefed
bluesky
bluesky-ai
17dAnthropic’s S-1 leaves two disclosure gaps, argue Ilan Strauss and Tim O’Reilly, co-directors of the AI Disclosures Project: its relationsh…Anthropic’s S-1 leaves two disclosure gaps, argue Ilan Strauss and Tim O’Reilly, co-directors of the AI Disclosures Project: its relat…
↗Not briefed
arxiv
arxiv-ai
17dMAC-v2: Long-horizon model-based value expansion from pixelsWe study whether \emph{long-horizon model-based value expansion} can serve as an effective framework in pixel-based offline reinforcem…
↗Not briefed
hn
hn-ai
17dShow HN: BrowserForge – AI agents use your real browser, you approve the clicks—
↗Not briefed
bluesky
bluesky-ai
17dMy pet theory is that claude code was invented by anthropic to sell more ClaudeMy pet theory is that claude code was invented by anthropic to sell more Claude
↗Not briefed
hn
hn-ai
17dPostal Service Surveilled Protest Activities Online with Little Rules, Oversight—
↗Not briefed
hn
hn-ai
17dIndependent Digital Media Is Not a Privilege Governments Can Revoke—
↗Not briefed
arxiv
arxiv-ai
17d$D^2$-OPD: Separating Direction and Distance in Extrapolative On-Policy DistillationOn-policy distillation (OPD) trains a student model to match a teacher's next-token distributions on student-generated prefixes. Extra…
↗Not briefed
arxiv
arxiv-ai
17dHow Do AI Agents Waste Your Money?AI agents may consume massive tokens across a wide range of tasks, but how many of those tokens are really necessary? Although prior w…
↗Not briefed
arxiv
arxiv-ai
17dArchetypal contexts in the neurons of language models: an architecture for intelligence?We validate our previous hypothesis that a language model holds relational patterns apart from the subject domains that instantiate th…
↗Not briefed
bluesky
bluesky-ai
17dit sucks that I can look at prose, video, or images and tell you not just whether they're AI but what model they areit sucks that I can look at prose, video, or images and tell you not just whether they're AI but what model they are
↗Not briefed
arxiv
arxiv-ai
17dNeural Compression over Equivalence Classes: Exact Gauges and Functional Rate-DistortionNeural networks admit many parameterizations of the same or nearly the same function, but deployment typically stores whichever parame…
↗Not briefed
bluesky
bluesky-ai
17d‘Anthropic’s plan includes the ability for insiders to start selling up to 10% of their shares on their first trading days. That will be a …‘Anthropic’s plan includes the ability for insiders to start selling up to 10% of their shares on their first trading days. That will…
↗Not briefed
arxiv
arxiv-ai
17dRecursive Sparse Attention: Million-Token Length Extrapolation at Logarithmic CostAt context lengths of about 1M tokens, dense attention is costly in prefill and decoding, and training typically relies on long docume…
↗Not briefed
hn
hn-ai
17dIntroducing Images 2.6—
↗Not briefed
bluesky
bluesky-ai
17di don’t ever use computer use, so i have no read on that but DSH (DeepSeek Harness) is well engineered, has a solid web UI, excellent plu…i don’t ever use computer use, so i have no read on that but DSH (DeepSeek Harness) is well engineered, has a solid web UI, excellen…
↗Not briefed
bluesky
bluesky-ai
17dInteresting exercise in AI alignment: I caught my 10 yo prompting Claude on my account before doing his homework. Whom do you think will Cl…Interesting exercise in AI alignment: I caught my 10 yo prompting Claude on my account before doing his homework. Whom do you think wi…
↗Not briefed
arxiv
arxiv-ai
17dSource-balanced Capped Fisher: Bounded Evidence Aggregation for User Inference in Language ModelsUser inference asks whether a source contributed documents to a language-model update, even when the auditor observes only fresh docum…
↗Not briefed
arxiv
arxiv-ai
17dLoopback Transformer: Towards a Unified View of Residual Looped TransformersLooped Transformers reduce parameter count by repeatedly applying shared blocks, but parameter sharing alone does not determine how th…
↗Not briefed
rss
techmeme
17dData center shares fall as investors expect slower AI buildout after Democrats win Congress; Hut 8 drops 23%, IREN 11%, Core Scientific 10%, Applied Digital 7% (Reuters)Reuters: Data center shares fall as investors expect slower AI buildout after Democrats win Congress; Hut 8 drops 23%, IREN 11%, Core…
↗Not briefed
bluesky
bluesky-ai
17dgithub.com/openai/codex... chabbt click reboot for me make no mistakegithub.com/openai/codex... chabbt click reboot for me make no mistake
↗Not briefed
bluesky
bluesky-ai
17dWorking with AI -- Vol.57 Some kind of dark lightning effect. I had DeepSeek add an alpha blend mode to the shader, and then showed it the …Working with AI -- Vol.57 Some kind of dark lightning effect. I had DeepSeek add an alpha blend mode to the shader, and then showed it…
↗Not briefed
bluesky
bluesky-ai
17danyone but schumer prospect.org/2026/09/11/o...anyone but schumer prospect.org/2026/09/11/o...
↗Not briefed
bluesky
bluesky-ai
17d'Super Mario Odyssey' Can Now Run on Apple TV Thanks to AI'Super Mario Odyssey' Can Now Run on Apple TV Thanks to AI
↗Not briefed
arxiv
arxiv-ai
17dSPILLWAY: Separating Target Admission from Sanitization to Defend Tool-Using Agents against Prompt InjectionTool-using agents act on email, files, and payments, so an instruction hidden in a web page or a shared document can make an agent sen…
↗Not briefed
bluesky
bluesky-ai
17danthropic would be unstoppable if they knew how to hire at least one engineer who gives a shit about anything other than sci-fi and hockeys…anthropic would be unstoppable if they knew how to hire at least one engineer who gives a shit about anything other than sci-fi and ho…
↗Not briefed
rss
wired-ai
17dThe Blue Wave Has Tech Bros Scrambling for Data Centers in SpaceAfter Democrats won big in Texas and other AI infrastructure hot spots on Tuesday, data center enthusiasts looked for comfort in low-E…
↗Not briefed
hn
hn-ai
17dMarkdown viewer that supports images and latex folmulas—
↗Not briefed
hn
hn-ai
17dNvidia store pulls RTX 5080 following RTX 5090 removal, prices rise to 2.2X MSRP—
↗Not briefed
arxiv
arxiv-ai
17dSAFE-Teen: Evaluating Chatbot Safety Across Adolescent Mental Health ConversationsAdolescents are increasingly turning to AI chatbots for mental health support, advice, and companionship, raising safety concerns. Whi…
↗Not briefed
arxiv
arxiv-ai
17dTileQuant: Tile as the Precision-Allocation Unit for LLM Mixed-Precision Post-Training QuantizationPost-training quantization (PTQ) reduces large language model (LLM) deployment costs, but preserving quality requires accounting for u…
↗Not briefed
arxiv
arxiv-ai
17dCompositional Unalignment: Auditing Safety in Large Vision-Language Models🔧 ModalLarge vision-language models (LVLMs) are vulnerable to compositional harm: an image and a prompt that are benign in isolation become h…
↗Not briefed
arxiv
arxiv-ai
17dFractional physics-informed neural networks to model social media viralityExisting models of social media virality either prescribe the functional form of engagement growth or assume a fixed mechanism by whic…
↗Not briefed
hn
hn-ai
17dGarlicSignage – open-source Digital Signage Software—
↗Not briefed
hn
hn-ai
17dBackground subagents in Claude Code now use less of your usage limit—
↗Not briefed