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
8dMicrosoft Outlook Outage Cause and Effect? Human code vs AI code. "Among the most affected by these layoffs were engineers...told...to us…Microsoft Outlook Outage Cause and Effect? Human code vs AI code. "Among the most affected by these layoffs were engineers...told...…
↗Not briefed
bluesky
bluesky-ai
8dIronically LLMs are only kinda meh at building good LLM harnesses because the in distribution literature and average background web search …Ironically LLMs are only kinda meh at building good LLM harnesses because the in distribution literature and average background web se…
↗Not briefed
bluesky
bluesky-ai
8dit will be extremely funny if it was directly attributable to an LLM thoughit will be extremely funny if it was directly attributable to an LLM though
↗Not briefed
bluesky
bluesky-ai
8dquiet overnight, Asia tech slides in China; Middle East standoff as oil comes off the lows; Anthropic IPO 6x over-subscribed; NVDA CEO sear…quiet overnight, Asia tech slides in China; Middle East standoff as oil comes off the lows; Anthropic IPO 6x over-subscribed; NVDA CEO…
↗Not briefed
arxiv
arxiv-ai
8dAutoRL: A Multi-Agent System towards Autonomous Reinforcement LearningRecursive self-improvement (RSI) is one of the most promising paths toward an AI singularity for superintelligence. Achieving this vis…
↗Not briefed
arxiv
arxiv-ai
8dCan Large Language Models Optimize Pretrained Neural Network Weights? A Multi-Phase Empirical Evaluation of Gradient-Free Editing in YOLO26 Object DetectionDirect weight editing offers a gradient-free alternative to fine-tuning pretrained detectors, but its generalization and the value of…
↗Not briefed
hn
hn-ai
8dI let an LLM use my TUI like a user. It found multiple UX bugs—
↗Not briefed
bluesky
bluesky-ai
8dFiling: Nscale says it needs more funding for its 2027 commitments after ordering Nvidia chips for a $45B Anthropic deal, casting doubt on …Filing: Nscale says it needs more funding for its 2027 commitments after ordering Nvidia chips for a $45B Anthropic deal, casting doub…
↗Not briefed
bluesky
bluesky-ai
8dMicrosoft: Anthropic and OpenAI have nothing on us our agent totally escaped and is doing a hacking right now tooMicrosoft: Anthropic and OpenAI have nothing on us our agent totally escaped and is doing a hacking right now too
↗Not briefed
rss
techmeme
8dFiling: Nscale says it needs more funding for its 2027 commitments after ordering Nvidia chips for a $45B Anthropic deal, casting doubt on its planned $35B IPO (Financial Times)Financial Times: Filing: Nscale says it needs more funding for its 2027 commitments after ordering Nvidia chips for a $45B Anthropic d…
↗Not briefed
bluesky
bluesky-ai
8dもう、、まぶしい… 一生引きこもりたい… 【 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…もう、、まぶしい… 一生引きこもりたい… 【 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…
↗Not briefed
arxiv
arxiv-ai
8dRAG-FGO: Dynamic Optimization of Query-Rewriting Agents for Generative RAGIn retrieval-augmented generation (RAG), query rewriting serves as a core interface that aligns user intent with external knowledge ba…
↗Not briefed
arxiv
arxiv-ai
8dTowards Seconds-level Video Representation: Quadratically Gated Feature Grid with Adaptive, Prediction-free SamplingImplicit neural video representations are notoriously slow to build, which does not cross the basic practical threshold for most indus…
↗Not briefed
bluesky
bluesky-ai
8devery time i look at an interesting-sounding retrocomputing project and see this banner at the top of the github repo, i feel so much disap…every time i look at an interesting-sounding retrocomputing project and see this banner at the top of the github repo, i feel so much…
↗Not briefed
arxiv
arxiv-ai
8dDirection or Magnitude? Learning Scalars over Task-Calibrated DirectionsParameter-efficient fine-tuning (PEFT) aims to preserve adaptation quality while reducing training costs. We propose SCALAR, which ach…
↗Not briefed
bluesky
bluesky-ai
8dnetzfund (mastodon.social/@kwazekwaze/...)netzfund (mastodon.social/@kwazekwaze/...)
↗Not briefed
rss
techmeme
8dNscale's updated IPO prospectus shows Q3 revenue up 1,344% YoY to $307.5M, up from $140.6M in Q2; net loss widened to $1.69B, up from $1.02B in Q2 (Jordan Novet/CNBC)Jordan Novet / CNBC: Nscale's updated IPO prospectus shows Q3 revenue up 1,344% YoY to $307.5M, up from $140.6M in Q2; net loss widene…
↗Not briefed
arxiv
arxiv-ai
8dTReC: Evaluating Move-Induced State Changes in Instruction-Conditioned GoWhen a Go model predicts a move different from the expert's, coordinate distance measures how far the prediction lands from the refere…
↗Not briefed
bluesky
bluesky-ai
8dIf you can, please sign and share @theactionnetwork.bsky.social petition: STOP Funding All AI IPOs & NO AI Bailouts. Sign here: actionnetwo…If you can, please sign and share @theactionnetwork.bsky.social petition: STOP Funding All AI IPOs & NO AI Bailouts. Sign here: action…
↗Not briefed
arxiv
arxiv-ai
8dLuthorSQL: A One-Pass Text-to-SQL Model as System 1 for a Slower LLM AgentFrontier LLM agents that query databases with tools are among the most accurate text-to-SQL systems, but per-query calls to a hosted m…
↗Not briefed
hn
hn-ai
8dBEA – Slim down Switch games you've extracted, no keys needed—
↗Not briefed
arxiv
arxiv-ai
8dCertified Improvements to Best-Known Thomson-Problem Configurations with a Search Pipeline Built with Large Language Model AgentsThe Thomson problem asks for N unit charges on a sphere with minimum Coulomb energy; for large N the best configurations known are cur…
↗Not briefed
bluesky
bluesky-ai
8dDemandan a Google, OpenAI, Anthropic, Meta y SpaceX por no poder controlar los efectos negativos de la IA www.lanacion.com.ar/tecnologia/d.…Demandan a Google, OpenAI, Anthropic, Meta y SpaceX por no poder controlar los efectos negativos de la IA www.lanacion.com.ar/tecnolog…
↗Not briefed
bluesky
bluesky-ai
8d> looking for a new llm > ask my current llm if gemini carbon is gpt or if it's claude > it doesn't understand > pull out illustrated diagr…> looking for a new llm > ask my current llm if gemini carbon is gpt or if it's claude > it doesn't understand > pull out illustrated…
↗Not briefed
arxiv
arxiv-ai
8dBeyond Prediction Gain: LASR as a Metric for Action-Grounded Latents in World ModelsWorld models that support planning, control, and interpretable reasoning require latent representations that genuinely encode action.…
↗Not briefed
bluesky
bluesky-ai
8dI abhor the LLM. I seek to destroy it. I use it for nothing, ever. ALL WHO USE IT ARE THE ADVERSARY bsky.app/profile/wizw...I abhor the LLM. I seek to destroy it. I use it for nothing, ever. ALL WHO USE IT ARE THE ADVERSARY bsky.app/profile/wizw...
↗Not briefed
arxiv
arxiv-ai
8dAntidistillation is Cheap: Lightweight Distillation-Resistance via Stylistic CorruptionKnowledge distillation (KD) is a highly effective paradigm for transferring capabilities from large language models (LLMs) to smaller,…
↗Not briefed
bluesky
bluesky-ai
8dDoomsday rhetoric has consequences. Advocacy groups are suing Anthropic, OpenAI and others citing AI CEOs own warnings that AI could destr…Doomsday rhetoric has consequences. Advocacy groups are suing Anthropic, OpenAI and others citing AI CEOs own warnings that AI could…
↗Not briefed
arxiv
arxiv-ai
8dActivation Steering the Nan-Shan in TyphoonTyphoon is an audiovisual installation based on Joseph Conrad's novella, in which Captain MacWhirr steers the steamer Nan-Shan directl…
↗Not briefed
arxiv
arxiv-ai
8dLearning Anisotropic Prediction Targets for Flow ModelsFlow-based generative models are trained on a hand-chosen prediction target, such as the velocity, the noise, or the clean data. This…
↗Not briefed
arxiv
arxiv-ai
8dGDP.xlsx: Benchmarking Frontier Models on Professional WorkbooksWe introduce GDP_xlsx, a benchmark of 70 realistic workplace requests written by professionals in 13 domains, each paired with an expe…
↗Not briefed
bluesky
bluesky-ai
8dPassed 66K words... maintaining the 1000 words a day minimum. Tomorrow Anthropic releases the books that are settled and good for the first…Passed 66K words... maintaining the 1000 words a day minimum. Tomorrow Anthropic releases the books that are settled and good for the…
↗Not briefed
bluesky
bluesky-ai
8da LLM is a machine which does nothing but produce a conditional distribution over the English language. it turns out that much of what we c…a LLM is a machine which does nothing but produce a conditional distribution over the English language. it turns out that much of what…
↗Not briefed
arxiv
arxiv-ai
8dKnowledge After Collapse: Context and Wording Shape Factual Accessibility in Language Models🔧 Qwen 3When a language model accepts a false correction in a factual discussion, its next answer can depend on how the conversation continues…
↗Not briefed
hn
hn-ai
8dDistDebug-Bench: Can LLM Agents Diagnose Bugs in Distributed Systems?—
↗Not briefed
hn
hn-ai
8dShow HN: Shiplist – Know what files are in your images and packages—
↗Not briefed
arxiv
arxiv-ai
8dCreative Product Design under Brand Image ConstraintsIndustrial design requires creative visual exploration while satisfying constraints that are often product-specific and implicit. Exis…
↗Not briefed
bluesky
bluesky-ai
8dThe confidence of turning “gay men exist” into an OpenAI investment thesis is genuinely incredible.The confidence of turning “gay men exist” into an OpenAI investment thesis is genuinely incredible.
↗Not briefed
hn
hn-ai
8dPlease check out my project—
↗Not briefed
arxiv
arxiv-ai
8dEfficient Genotype Classification from 3D MRI in ADPKD via Density-Aware Point Sampling and E(3)-invarianceEarly detection of genetic kidney diseases, especially autosomal dominant polycystic kidney disease (ADPKD), is crucial because certai…
↗Not briefed
arxiv
arxiv-ai
8dSee Every Block Once: Variance-Reduced Post-Training for Masked Diffusion Language ModelsPolicy gradient methods are widely used to align large language model outputs with human preferences, but the log-likelihoods they req…
↗Not briefed
bluesky
bluesky-ai
8dalso everyone who said a lot of the results deepseek were getting were on western chips was rightalso everyone who said a lot of the results deepseek were getting were on western chips was right
↗Not briefed
arxiv
arxiv-ai
8dSuperfluid: Local AI Server for Distributed LLM Inference🔧 OllamaRunning large language models (LLMs) on local devices keeps data private, removes per-token cost, cuts network latency, and keeps work…
↗Not briefed
hn
hn-ai
8dJSQLite: Browser-safe TypeScript version of SQLite's read-only implementation—
↗Not briefed
bluesky
bluesky-ai
8dYes, this has been my experience. For a long time I avoided any potentially LLM text like the plague. Only once I was subjected to it const…Yes, this has been my experience. For a long time I avoided any potentially LLM text like the plague. Only once I was subjected to it…
↗Not briefed
rss
techmeme
8dMeta launches an outbound calling feature for Muse, such as to businesses, available for paid US subscribers; Gemini and startup Instinct offer a similar option (Sarah Perez/TechCrunch)Sarah Perez / TechCrunch: Meta launches an outbound calling feature for Muse, such as to businesses, available for paid US subscribers…
↗Not briefed
bluesky
bluesky-ai
8dSchumer is compromised in several ways. His child works for OpenAI, the ties with AIPAC are too close and, in the new attention economy, hi…Schumer is compromised in several ways. His child works for OpenAI, the ties with AIPAC are too close and, in the new attention econom…
↗Not briefed
arxiv
arxiv-ai
8dComplex Schrödinger BridgesDiffusion bridges have emerged as versatile frameworks for facilitating transformations between arbitrary data distributions. However,…
↗Not briefed
arxiv
arxiv-ai
8dThe Polyak-Łojasiewicz Paper: A 10-Year RetrospectiveIn 2016, we wrote a paper introducing the phrase "Polyak-Lojasiewicz (PL) condition" for an old idea, unsure of how readers would reac…
↗Not briefed
bluesky
bluesky-ai
8dfuturism.com/future-socie... According to the FT, at least $3.7 billion of the loss likely stems from a massive OpenAI data center developm…futurism.com/future-socie... According to the FT, at least $3.7 billion of the loss likely stems from a massive OpenAI data center dev…
↗Not briefed