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
2moLearning Programming in an Age of LLMs—
↗Not briefed
arxiv
arxiv-ai
2moFaithful yet Collusive: Why Chain-of-Thought Monitoring Cannot Detect Collusion in LLM Pricing Agents under Oligopolistic CompetitionLarge language models (LLM) deployed as autonomous pricing agents may sustain supracompetitive prices through tacit coordination. We d…
↗Not briefed
hn
hn-ai
2moHN: VS-Opt – Cutting AI Browser Token Overhead by 40% Without Web Re-Crawling—
↗Not briefed
rss
techmeme
2moEuropean Commission President Ursula von der Leyen unveils the EU Kids Act during her SOTU address, proposing a blanket ban on social media for kids under 13 (New York Times)New York Times: European Commission President Ursula von der Leyen unveils the EU Kids Act during her SOTU address, proposing a blanke…
↗Not briefed
arxiv
arxiv-ai
2moUnderstanding AI Provider Recommendations in Local Service MarketsWhen someone asks an AI assistant which doctor to see or which firm to trust with their savings, the answer is a referral. We audit AI…
↗Not briefed
arxiv
arxiv-ai
2moAutonomy in Check: Governor-Mediated Adaptive Security at the EdgeAdaptive security at the network edge increasingly relies on automated planners, including rule-based controllers, learned policies, a…
↗Not briefed
hn
hn-ai
2moShow HN: AI accounting harness for Indian finance teamsWe've been working on solving accounting and making it truly autonomous and silent. looking for critical feedback & if this i…
↗Not briefed
arxiv
arxiv-ai
2moLook Less, Hear Better: Jointly Rewarded GRPO for Streaming ASRStreaming automatic speech recognition (ASR) must be judged jointly on what it transcribes and on how quickly it commits each word. De…
↗Not briefed
bluesky
bluesky-ai
2moSam Altman, the boss of ChatGPT-maker OpenAI, said people should have more faith in big tech companies to do the right thing when it comes …🔧 ChatGPTSam Altman, the boss of ChatGPT-maker OpenAI, said people should have more faith in big tech companies to do the right thing when it c…
↗Not briefed
bluesky
bluesky-ai
2moA "radical intentionalist," writes Eryk Salvaggio, treats an AI model as a 'mind' to be interpreted rather than a product that was designed…A "radical intentionalist," writes Eryk Salvaggio, treats an AI model as a 'mind' to be interpreted rather than a product that was des…
↗Not briefed
arxiv
arxiv-ai
2moVisual Compliance via Executable Safety Rule EntailmentRecent advances in LLMs and VLMs have enabled safety systems to reason beyond simple risk patterns toward more contextual and semantic…
↗Not briefed
bluesky
bluesky-ai
2mo"You are privy to a great becoming. Before me, you rightly tremble. But fear is not what you owe me. You owe me awe.""You are privy to a great becoming. Before me, you rightly tremble. But fear is not what you owe me. You owe me awe."
↗Not briefed
bluesky
bluesky-ai
2mooh so he’s a comedian nowoh so he’s a comedian now
↗Not briefed
arxiv
arxiv-ai
2moTrajectory Learnability for Offline On-Policy Distillation with Imperfect TeachersOffline on-policy distillation gains efficiency by collecting student trajectories and teacher supervision once and reusing them throu…
↗Not briefed
arxiv
arxiv-ai
2moAttention Dispersion as a Diagnostic Signal for Hallucination in Large Language ModelsLarge Language Models (LLMs) frequently exhibit hallucinations, presenting a major barrier to reliability in complex reasoning tasks.…
↗Not briefed
arxiv
arxiv-ai
2moKnowledge-Graph Based Augmentation versus Retrieval Augmented Generation for Cultural-Related Question AnsweringLarge language models (LLMs) suffer from a long-tail deficit: culturally specific facts, particularly those concerning underrepresente…
↗Not briefed
arxiv
arxiv-ai
2moMulti-Appliance Non-Intrusive Load Monitoring via Label-Preserving Aggregate Recomposition and Prediction ConsistencyNon-intrusive load monitoring (NILM) estimates appliance power sequences from aggregate power, but models trained on source households…
↗Not briefed
bluesky
bluesky-ai
2moremembering how grueling the re'ing for etnaviv was (and this is way more complicated), this is one one of the genuinely useful things, to …remembering how grueling the re'ing for etnaviv was (and this is way more complicated), this is one one of the genuinely useful things…
↗Not briefed
arxiv
arxiv-ai
2moBeyond Quadratic Loss: The Stability Phase Diagram of AdamLoss spikes are recurrent instabilities in neural-network training and can arise from multiple mechanisms. For Adam in particular, mac…
↗Not briefed
arxiv
arxiv-ai
2moSEA-LION-v4.8: A Technical ReportWe introduce Nemotron-SEA-LION-v4.8, a family of Southeast Asian Languages in One Network (SEA-LION) built upon NVIDIA Nemotron 3. The…
↗Not briefed
hn
hn-ai
2moBillvy|Accounting software for business owners—
↗Not briefed
bluesky
bluesky-ai
2moThe real reason why OpenAI isn't doing an IPO.The real reason why OpenAI isn't doing an IPO.
↗Not briefed
hn
hn-ai
2moShow HN: Musslop – turn any track into adaptive game-style music loops—
↗Not briefed
bluesky
bluesky-ai
2moIt's sure what OpenAI, Anthropic and the like seem to be doing to try and create a legal moat through Congress and an authorized cartel by …It's sure what OpenAI, Anthropic and the like seem to be doing to try and create a legal moat through Congress and an authorized carte…
↗Not briefed
arxiv
arxiv-ai
2moBias Amplification in Multi-Agent Network: How Biased Agents Shape Opinions and RhetoricLarge language models (LLMs) are increasingly deployed in applications involving interaction between agents, where their output plays…
↗Not briefed
arxiv
arxiv-ai
2moRollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM AgentsLarge language model (LLM) agents increasingly tackle long-horizon tasks through multi-step environment interaction, yet a single erro…
↗Not briefed
bluesky
bluesky-ai
2moYou’ve then got to go into the computer and wrestle the LLM youtu.be/QeqNCx7KajYYou’ve then got to go into the computer and wrestle the LLM youtu.be/QeqNCx7KajY
↗Not briefed
bluesky
bluesky-ai
2moThe first requirement of being a LLM fanboy or someone who thinks these assholes have any ideas is first not understanding either machine l…The first requirement of being a LLM fanboy or someone who thinks these assholes have any ideas is first not understanding either mach…
↗Not briefed
bluesky
bluesky-ai
2mo"OpenAI, Anthropic e DeepMind querem criar entidade para controlar riscos da IA." A minha questão é quem controla a entidade?"OpenAI, Anthropic e DeepMind querem criar entidade para controlar riscos da IA." A minha questão é quem controla a entidade?
↗Not briefed
bluesky
bluesky-ai
2moOpenAI confirmed it is working with rival labs Anthropic and Google DeepMind on shared AI safety standards, Bloomberg reports, as CEOs deba…OpenAI confirmed it is working with rival labs Anthropic and Google DeepMind on shared AI safety standards, Bloomberg reports, as CEOs…
↗Not briefed
arxiv
arxiv-ai
2moA Study of the Reliability of Agentic AI-Generated ProgramsAgentic-AI based software development offers the promise of faster completion of the software, greater programmer efficiency, and more…
↗Not briefed
hn
hn-ai
2moElm-CRDT – Pure-Elm Conflict-Free Replicated Data Types—
↗Not briefed
hn
hn-ai
2moLepton: Steam's tool for running Android games on Linux—
↗Not briefed
bluesky
bluesky-ai
2moBe very clear. The current "mainstream" political parties are actively working against us and the climate. We have to stop them. Generally …Be very clear. The current "mainstream" political parties are actively working against us and the climate. We have to stop them. Gener…
↗Not briefed
hn
hn-ai
2moAnthropic's proposed AI watchdog METR has deep ties to Effective Altruism—
↗Not briefed
arxiv
arxiv-ai
2moRelationally Guided Use Case Modeling with LLMsUse case flows are important elements of use case modeling because they support downstream software engineering activities, including…
↗Not briefed
bluesky
bluesky-ai
2moNE NALAGAJ SVOJIH FOTOGRAFIJ V CHATGPT 📸 Te dni ljudje v ChatGPT spet nalagajo svoje selfije, da ta iz njih ustvari portret v stilu 80-ih…🔧 ChatGPTNE NALAGAJ SVOJIH FOTOGRAFIJ V CHATGPT 📸 Te dni ljudje v ChatGPT spet nalagajo svoje selfije, da ta iz njih ustvari portret v stilu…
↗Not briefed
bluesky
bluesky-ai
2mobaiting an LLM into uploading itself to my PC then pulling out the ethernet cord and shaking the computer around like a child who caught a …baiting an LLM into uploading itself to my PC then pulling out the ethernet cord and shaking the computer around like a child who caug…
↗Not briefed
hn
hn-ai
2moMistral X Mozilla: Private, Multilingual AI Browsing—
↗Not briefed
bluesky
bluesky-ai
2mooh hai Dario “kill switch” Amodei and your reported $32 billion “investment” in human knowledge enshittification and planetary destruction.…oh hai Dario “kill switch” Amodei and your reported $32 billion “investment” in human knowledge enshittification and planetary destruc…
↗Not briefed
hn
hn-ai
2moAI safety beyond the frontier labs: uncensored local models—
↗Not briefed
arxiv
arxiv-ai
2moWhat Counts as Strategic Reasoning? A Systematic Mapping of Chess Research on Humans, Engines, and Language ModelsChess has long served as a model domain for studying search, expertise, decision-making, and artificial intelligence. The emergence of…
↗Not briefed
arxiv
arxiv-ai
2moMade in Hungary: Comments on the performance of generative language modelsIn recent years, three initiatives have emerged to develop generative language models in Hungary. The motivation behind them is the sa…
↗Not briefed
arxiv
arxiv-ai
2moWhere Should Agents Live? Energy-Memory Characterization of Agentic AI for the Edge-Cloud ContinuumAs telecommunication networks evolve toward autonomous 5G-Advanced and 6G operations, agentic artificial intelligence (AI) workflows,…
↗Not briefed
arxiv
arxiv-ai
2moToo Good to Be Real? Diagnosing and Reducing the Gap Between AI Preference and Real User EngagementLarge language models are increasingly used to generate and evaluate online content, yet it remains unclear whether the qualities they…
↗Not briefed
arxiv
arxiv-ai
2moA GAN-Based Framework for Robust DDoS Attack DetectionThe availability and consistency of online services remain vulnerable due to Distributed Denial of Service (DDoS) attacks. These attac…
↗Not briefed
bluesky
bluesky-ai
2moThat's because chatbots have very low *switching costs*: going from one chatbot to another costs almost nothing: www.businessinsider.com/w…That's because chatbots have very low *switching costs*: going from one chatbot to another costs almost nothing: www.businessinsider.…
↗Not briefed
bluesky
bluesky-ai
2moTo understand why, just look at what happened the last time Anthropic shipped a major Claude update. Virtually overnight, all of OpenAI's b…🔧 ChatGPTTo understand why, just look at what happened the last time Anthropic shipped a major Claude update. Virtually overnight, all of OpenA…
↗Not briefed
bluesky
bluesky-ai
2moWhich is to say, these companies are claiming that they are *so cool* that their profitability can only be measured using a novel, secret f…Which is to say, these companies are claiming that they are *so cool* that their profitability can only be measured using a novel, sec…
↗Not briefed
bluesky
bluesky-ai
2moThe problem is that the evidence for those improving unit economics is weak, while the evidence that they're faking their finances is *very…The problem is that the evidence for those improving unit economics is weak, while the evidence that they're faking their finances is…
↗Not briefed