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
1moShow HN: Browse last quarter's Supreme Court cases, linked and cited as you readI made this as a means to browse through documents that are highly linked and referential in general, and thought it would be interest…
↗Not briefed
arxiv
arxiv-ai
1moCorrelation-Free Transition Path Sampling through Shooting Point Generation Guided by Committor LearningStudying the dynamical behavior of a system often depends on characterizing how it transitions between long-lived states. Because such…
↗Not briefed
bluesky
bluesky-ai
1moMustafa Suleyman on the AI safety debate, why AI isn't consciousness, and how the industry pushes toward alignment. www.theverge.com/podcas…Mustafa Suleyman on the AI safety debate, why AI isn't consciousness, and how the industry pushes toward alignment. www.theverge.com/p…
↗Not briefed
hn
hn-ai
1moLawsuit may lead to lowered healthcare plan standards for employees in the USA—
↗Not briefed
hn
hn-ai
1moShow HN: Unimodal – Our AI ManifestWe&#x27;re launching some products that need to get built to enable what&#x27;s coming next.<p>The focus is around AI economies and Hu…
↗Not briefed
bluesky
bluesky-ai
1moAbsolutely love that the entirety of the AI and LLM related discourse is all about a weird alleged sex cult now. Great job, y'all!Absolutely love that the entirety of the AI and LLM related discourse is all about a weird alleged sex cult now. Great job, y'all!
↗Not briefed
bluesky
bluesky-ai
1mo2022: A PoC of an LLM research demo breaks containment and sees widespread adoption SV:2022: A PoC of an LLM research demo breaks containment and sees widespread adoption SV:
↗Not briefed
bluesky
bluesky-ai
1mo😞 open.substack.com/pub/dropsite...😞 open.substack.com/pub/dropsite...
↗Not briefed
arxiv
arxiv-ai
1moFingerprinting Multimodal Large Language Models🔧 ModalWhile multimodal large language models (MLLMs) enable a wide range of image-text reasoning tasks, recent incidents indicate that they…
↗Not briefed
arxiv
arxiv-ai
1moSkillAA: Attribution-Guided Skill-Graph Updating with Targeted Validation and Rollback🔧 GPT-5External skills provide domain procedures without parameter updates, but existing methods often edit skills directly from failed rollo…
↗Not briefed
bluesky
bluesky-ai
1moI feel like Metamorphic testing, checking how changing an input changes a function, would be a pretty good way to evaluate AI tooling. Lik…I feel like Metamorphic testing, checking how changing an input changes a function, would be a pretty good way to evaluate AI tooling.…
↗Not briefed
arxiv
arxiv-ai
1moOnline Supervised Dimension Reduction with Random Features: Diagnostics and Computational Trade-offsAccurate optimization of a supervised spectral objective need not produce an accurate population subspace or a better predictive repre…
↗Not briefed
hn
hn-ai
1moShow HN: Session-trail – A timeline of your project, not a graph of your code—
↗Not briefed
arxiv
arxiv-ai
1moSeismic Site Response Prediction from Sparse Observations Using Finite-Element-Pretrained Latent DynamicsNumerical site-response predictions often deviate from observations, yet correcting these discrepancies is difficult because records a…
↗Not briefed
hn
hn-ai
1moGraphQL-backed MCP tools are more token-efficient—
↗Not briefed
bluesky
bluesky-ai
1moContinues #palworld on my server & binge some #hellskitchen #openai #chatgpt #Xai #gemini #ghostinthemachine #angels #void #haven #roguene…🔧 ChatGPTContinues #palworld on my server & binge some #hellskitchen #openai #chatgpt #Xai #gemini #ghostinthemachine #angels #void #haven #ro…
↗Not briefed
arxiv
arxiv-ai
1moThe Organization of Inference: Information, Resource Constraints, and AI ProductionThe economic value of inference depends on how capacity and task information are distributed across stages of AI production. We study…
↗Not briefed
bluesky
bluesky-ai
1moi also think that pricing is optimistic considering currently ~$200 p/m of tokens has an actual cost of ~$2000. the moment they actually p…i also think that pricing is optimistic considering currently ~$200 p/m of tokens has an actual cost of ~$2000. the moment they actua…
↗Not briefed
bluesky
bluesky-ai
1moTruly astonishing that all any world leader can think to do is keep inviting the same guys into a room to give their considered advice on h…Truly astonishing that all any world leader can think to do is keep inviting the same guys into a room to give their considered advice…
↗Not briefed
arxiv
arxiv-ai
1moCross-Architecture Foundation-Model Distillation for Edge Flood SegmentationGeospatial foundation models can provide strong flood-segmentation performance, but their size limits deployment on memory-constrained…
↗Not briefed
arxiv
arxiv-ai
1moA Mathematical Model of Motivated Emotional Mind - Cognitive Embodied SystemThis article presents a mathematical model of the Motivated Emotional Mind cognitive architecture developed for embodied intelligent s…
↗Not briefed
bluesky
bluesky-ai
1moAI makes it so easy to cheat, it's decided to do a little cheating of its own. www.washingtonpost.com/technology/2...AI makes it so easy to cheat, it's decided to do a little cheating of its own. www.washingtonpost.com/technology/2...
↗Not briefed
bluesky
bluesky-ai
1moOpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior www.nytimes.com/2026/09/16/t...OpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior www.nytimes.com/2026/09/16/t...
↗Not briefed
bluesky
bluesky-ai
1moOpenAI hat sechs neue Fälle von KI-Fehlverhalten gemeldet. Die KIs setzten sich über Nutzer hinweg, auch bei Alltagsaufgaben. #OpenAIOpenAI hat sechs neue Fälle von KI-Fehlverhalten gemeldet. Die KIs setzten sich über Nutzer hinweg, auch bei Alltagsaufgaben. #OpenAI
↗Not briefed
bluesky
bluesky-ai
1moOpenAI revealed six new concerning incidents involving its models.OpenAI revealed six new concerning incidents involving its models.
↗Not briefed
bluesky
bluesky-ai
1moNEW —> AI giants tap into the social media playbook: self-regulation Execs at Anthropic, OpenAI, Meta and Google are converging on industr…NEW —> AI giants tap into the social media playbook: self-regulation Execs at Anthropic, OpenAI, Meta and Google are converging on in…
↗Not briefed
arxiv
arxiv-ai
1moWhen Do Language-Grounded Explanations Help? A Graph-Bottleneck for Farm Monitoring Interpretable Sheep Facial PainAutomated pain recognition from facial expression could make continuous welfare assessment practical in sheep, but adoption depends on…
↗Not briefed
hn
hn-ai
1moOpenAI discloses new 'concerning' model behaviour—
↗Not briefed
rss
techcrunch-ai
1moHuawei plans Q1 2027 launch of new AI chip as it takes on NvidiaHuawei is accelerating the launch of its next-generation Ascend 960DT AI chip as it pushes to compete with Nvidia and close China’s AI…
↗Not briefed
rss
techmeme
1moKing Charles III hosts tech leaders, including Jensen Huang, Demis Hassabis, and OpenAI CFO Sarah Friar, to discuss AI risks; Huang calls for AI safety tests (Bloomberg)Bloomberg: King Charles III hosts tech leaders, including Jensen Huang, Demis Hassabis, and OpenAI CFO Sarah Friar, to discuss AI risk…
↗Not briefed
bluesky
bluesky-ai
1moMicrosoft AI CEO Mustafa Suleyman says AI threats are real, and Anthropic is making it worse www.theverge.com/podcast/9964...Microsoft AI CEO Mustafa Suleyman says AI threats are real, and Anthropic is making it worse www.theverge.com/podcast/9964...
↗Not briefed
hn
hn-ai
1moSwitch to Codex seamlessly when Claude Code is used up—
↗Not briefed
hn
hn-ai
1moMainline – A light macOS menu bar app to keep you on top of your GitHub PRs—
↗Not briefed
bluesky
bluesky-ai
1moIt’s really disturbing how much AI users lie about their use. The wiki noticeboard dealing with LLM use sees a commmon pattern: an editor a…It’s really disturbing how much AI users lie about their use. The wiki noticeboard dealing with LLM use sees a commmon pattern: an edi…
↗Not briefed
bluesky
bluesky-ai
1moHello, friends! I'm launching a new project: secret.systems I want to share my photography in a way that feels direct, personal, and real.…Hello, friends! I'm launching a new project: secret.systems I want to share my photography in a way that feels direct, personal, and…
↗Not briefed
arxiv
arxiv-ai
1moSCGFM-ART: Amortized Relational Transport for Structure-Centric Graph Foundation ModelsGraph foundation models (GFMs) aim to learn transferable representations across severely heterogeneous graph domains. However, severe…
↗Not briefed
bluesky
bluesky-ai
1moIf you need to fine-tune a model for your codebase, your code is bad. Agree or disagree? We talk AI hot takes in our latest GitHub Podca…If you need to fine-tune a model for your codebase, your code is bad. Agree or disagree? We talk AI hot takes in our latest GitHub…
↗Not briefed
bluesky
bluesky-ai
1moI hope we can all see how the marketing rhetoric tech companies use gets adopted by the press in a way that anthropomorphizes software and …I hope we can all see how the marketing rhetoric tech companies use gets adopted by the press in a way that anthropomorphizes software…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI discloses six new incidents of "unexpected or concerning" behavior by its AI models, and also unveils a new framework for tracking a…OpenAI discloses six new incidents of "unexpected or concerning" behavior by its AI models, and also unveils a new framework for track…
↗Not briefed
rss
techcrunch-ai
1mo2 days left to exhibit at TechCrunch Disrupt 2026Last day to book your exhibit table at Disrupt is September 18. Two days left. Get your startup in front of 10,000+ founders, investor…
↗Not briefed
rss
theverge-ai
1moMicrosoft AI CEO says AI threats are real, and Anthropic is making it worse🔧 Consensus🔧 Hugging FaceToday, I’m talking with Mustafa Suleyman, the CEO of Microsoft AI. As you’re no doubt aware, the biggest story in tech right now is th…
↗Not briefed
rss
theverge-ai
1moAI is feared globally as the destroyer of jobsIn 34 of the 37 surveyed countries, people are more likely to believe AI will lead to job losses over the next 20 years. | Image: Pew…
↗Not briefed
bluesky
bluesky-ai
1mosimilarly, it's not "model misalignment" it's automated hacking software (built and managed by incompetent, unethical people) doing exactly…similarly, it's not "model misalignment" it's automated hacking software (built and managed by incompetent, unethical people) doing ex…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI also said Wednesday it was introducing a new framework for tracking, probing and disclosing instances of what it called "misalignmen…OpenAI also said Wednesday it was introducing a new framework for tracking, probing and disclosing instances of what it called "misali…
↗Not briefed
arxiv
arxiv-ai
1moTouchSight: Bare-Handed Tactile Prediction from Egocentric Video via Generative Visual AugmentationTactile signals provide direct contact and force measurements that are essential for understanding physical interactions and enabling…
↗Not briefed
bluesky
bluesky-ai
1moit is not "acting out" it is doing exactly what it's being programmed to do, and the failures come because it's being overseen by incompete…it is not "acting out" it is doing exactly what it's being programmed to do, and the failures come because it's being overseen by inco…
↗Not briefed
hn
hn-ai
1moI Don't Like LLMs—
↗Not briefed
hn
hn-ai
1moAsk HN: How about small technological solutions to bring oil prices back down?The prices on diesel are approaching ludicrous. In my mind, this should be an opportunity for a new technological solution to be imple…
↗Not briefed
hn
hn-ai
1moShow HN: Dnswer – browser-based DNS change approval and multi-provider syncImagine you have to manage DNS for domains spread across multiple DNS providers. You want to apply consistent DNS settings across your…
↗Not briefed
arxiv
arxiv-ai
1moStress-testing Alignment MidtrainingWhen aligning frontier models through post-training techniques, it is not possible to directly demonstrate all of the behaviours we wa…
↗Not briefed