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
1moReviewBench: An open benchmark for AI code review—
↗Not briefed
bluesky
bluesky-ai
1moNota bene: his punchline is way better! The LLM pulled the punch.Nota bene: his punchline is way better! The LLM pulled the punch.
↗Not briefed
hn
hn-ai
1moBuying an Antminer Z15 Pro miner – 860K vs. 840K—
↗Not briefed
bluesky
bluesky-ai
1moSome basic trouble even discussing this: LLM theft-machine ('AI') proponents are self-selected for delusionality and intent to commit felon…Some basic trouble even discussing this: LLM theft-machine ('AI') proponents are self-selected for delusionality and intent to commit…
↗Not briefed
bluesky
bluesky-ai
1moFWIW this is the most ambitious benchmark I've co-created! 𝗞𝗡𝗢𝗪𝗦 jointly & reliably evaluates agents on: 1) live web search, 2) us…FWIW this is the most ambitious benchmark I've co-created! 𝗞𝗡𝗢𝗪𝗦 jointly & reliably evaluates agents on: 1) live web search,…
↗Not briefed
hn
hn-ai
1moSchneider Electric to buy US software firm PTC in $22.6B deal—
↗Not briefed
bluesky
bluesky-ai
1moThere is still not a single good frictionless consumer-facing AI app with a real-world use case I'm aware of. Can someone more plugged in …There is still not a single good frictionless consumer-facing AI app with a real-world use case I'm aware of. Can someone more plugge…
↗Not briefed
arxiv
arxiv-ai
1moDomain adaptation of Russian ModernBERT for long legal documentsWe investigate whether continued pretraining on Russian legislative documents improves a Russian ModernBERT encoder on legal text. The…
↗Not briefed
bluesky
bluesky-ai
1moAmong the many issues: If you learn something, act on it, and cause harm, you can be held liable for the results. If an LLM "learns" some…Among the many issues: If you learn something, act on it, and cause harm, you can be held liable for the results. If an LLM "learns"…
↗Not briefed
bluesky
bluesky-ai
1moone of my brother's friends ended up working at anthropic (not by choice; she worked at a different company they acquired) and her feelings…one of my brother's friends ended up working at anthropic (not by choice; she worked at a different company they acquired) and her fee…
↗Not briefed
bluesky
bluesky-ai
1moEverything is tiring because it's all dumb but it takes a lot of time to explain why it's dumb. Like you have to say shit like "Yeah the CE…Everything is tiring because it's all dumb but it takes a lot of time to explain why it's dumb. Like you have to say shit like "Yeah t…
↗Not briefed
hn
hn-ai
1moMetagente, a tiny language for AI agents that speaks MCP and A2A—
↗Not briefed
arxiv
arxiv-ai
1moOut-of-control Hamiltonian LearningLearning the Hamiltonian of a many-body system from its dynamics is a central task in quantum science, yet the algorithms with the str…
↗Not briefed
arxiv
arxiv-ai
1moSAFE-MR: Evidence Sufficiency Learning for Selective Multimodal Rumor DetectionMultimodal rumor detectors increasingly rely on retrieved evidence, yet relevant evidence is not necessarily sufficient for verificati…
↗Not briefed
hn
hn-ai
1moShow HN: Jiwo - Small decision models topping decision index leaderboard🔧 Qwen 3Spent the weekend turning small LLMs into decision models, with some good results and a lot of learnings along the way. Topping the de…
↗Not briefed
bluesky
bluesky-ai
1moAI was always a jobs program for CS Majors, also we need to separate the technology from the politics of how it's used somehow. Maybe that…AI was always a jobs program for CS Majors, also we need to separate the technology from the politics of how it's used somehow. Maybe…
↗Not briefed
arxiv
arxiv-ai
1moReading the Mood: Emotion-Guided Book-to-Music Recommendation via CGANs and LLMsBackground music that matches the mood of a text has been shown to make readers feel more immersed and improve their reading experienc…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic is now worth $965B. JPMorgan, the biggest bank in America, is worth $900B. One took 227 years to get here. The other took 5.Anthropic is now worth $965B. JPMorgan, the biggest bank in America, is worth $900B. One took 227 years to get here. The other took 5.
↗Not briefed
arxiv
arxiv-ai
1moA Solvable Model of Adaptive Learning Rate Rescaling: Acceleration, Stability & ScalingA recurring design principle in modern optimizers is to decouple update magnitude from the raw gradient norm, yet its consequences for…
↗Not briefed
hn
hn-ai
1moBenchmark in Milliseconds—
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
rss
bing-news-ai
1moDonald Trump Signs ‘Super Intelligence’ Agreement With Awkward MisspellingThe document includes the signatures of Google CEO Sundar Pichai, Anthropic co-founder Dario Amodei, Meta CEO Mark Zuckerberg, OpenAI…
↗Not briefed
rss
bing-news-ai
1moMusk Adopts Trump's 'Super Intelligence' Term for SpaceX AI UnitElon Musk said SpaceX will rename its artificial intelligence unit from SpaceXAI to SpaceXSI, adopting President Donald Trump's push t…
↗Not briefed
bluesky
bluesky-ai
1mo‘These are some of the hardest questions that we face or that anybody developing this technology faces,’ Altman said. www.theverge.com/ai-a…‘These are some of the hardest questions that we face or that anybody developing this technology faces,’ Altman said. www.theverge.com…
↗Not briefed
bluesky
bluesky-ai
1moDebian kernel update: 1,313 new CVEs Debian's latest Linux kernel security advisory for kernel 6.12.111-1 reports 1,313 CVEs, a tally like…Debian kernel update: 1,313 new CVEs Debian's latest Linux kernel security advisory for kernel 6.12.111-1 reports 1,313 CVEs, a tally…
↗Not briefed
hn
hn-ai
1moShow HN: Self-bench – benchmark coding agents on real-world softwareHey HN,<p>Today, we&#x27;re launching selfbench.dev, an open-source tool that lets you create and run evals automatically from your PR…
↗Not briefed
bluesky
bluesky-ai
1moOpen AI is flailing. Menin says it "stretches credulity" that OpenAI is committed to regulation when Altman & wife each gave $12.5M to a PA…Open AI is flailing. Menin says it "stretches credulity" that OpenAI is committed to regulation when Altman & wife each gave $12.5M to…
↗Not briefed
arxiv
arxiv-ai
1moMedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA TasksWhile medical multimodal large language models (Med-MLLMs) advance medical visual question answering (VQA), existing clinical workflow…
↗Not briefed
rss
theverge-ai
1moOpenAI PR tells journalist to ‘move on’ while asking Sam Altman about a ChatGPT user&#8217;s suicide🔧 ChatGPTAn OpenAI publicist tried to change the topic of CEO Sam Altman's interview with Vanity Fair's Mark Guiducci after the editor brought…
↗Not briefed
arxiv
arxiv-ai
1moTo Learn is to Wander: Learning Across Graphs and Tasks with Random WalksGraph foundation models aim to transfer across graphs, feature spaces, relational schemas, and prediction tasks, yet existing approach…
↗Not briefed
arxiv
arxiv-ai
1moAdapting prior-data fitted networks for tabular anomaly detectionWhile deep features have transformed anomaly detection in images and video, their impact on tabular data has been less substantial, pa…
↗Not briefed
bluesky
bluesky-ai
1moSam Altman: "Creemos que el mundo debería aceptar que ocurran algunas cosas negativas a cambio de los beneficios que aporta esta tecnología…Sam Altman: "Creemos que el mundo debería aceptar que ocurran algunas cosas negativas a cambio de los beneficios que aporta esta tecno…
↗Not briefed
arxiv
arxiv-ai
1moRevisiting Label-Free Speaker Embedding Enhancement with vMF Profile LikelihoodEmbedding enhancement improves speaker verification under acoustic mismatch without modifying a frozen backbone. Recent work has estab…
↗Not briefed
hn
hn-ai
1moLocalhost: Our First Conference and Why the Cloud Needs to Get Out of Your Way—
↗Not briefed
bluesky
bluesky-ai
1moOpenAI should be held legally accountable for any actions carried out by its AI. As a corporation and its individual officers. Then they …OpenAI should be held legally accountable for any actions carried out by its AI. As a corporation and its individual officers. Then…
↗Not briefed
arxiv
arxiv-ai
1moProgrammatic Search Agents: Extending Agentic Search Beyond Query ReformulationSearch agents adapt their queries, yet fixed search interfaces leave candidate processing and evidence presentation outside the agent'…
↗Not briefed
bluesky
bluesky-ai
1moThe biggest difference is *the LLM is not a person*. And if it IS a person? If it has a mind? That takes what these AI companies are doing…The biggest difference is *the LLM is not a person*. And if it IS a person? If it has a mind? That takes what these AI companies are…
↗Not briefed
arxiv
arxiv-ai
1moOVAL: Output-Aware Local Page Bases for KV Cache RetrievalLong context inference with large language models becomes increasingly expensive as attention must operate over an ever growing KV cac…
↗Not briefed
bluesky
bluesky-ai
1moProbably a good sign for the future of AI that they're using the "heroin dealer" model of attracting customersProbably a good sign for the future of AI that they're using the "heroin dealer" model of attracting customers
↗Not briefed
arxiv
arxiv-ai
1moAligning Multimodal Patient Evidence with Biomedical Knowledge Graphs for Clinical LLMsClinical questions often depend on linking a patient's multimodal evidence to external biomedical knowledge, yet existing predictive s…
↗Not briefed
bluesky
bluesky-ai
1moBecause an LLM doesn't "learn" from training materials, it (vastly simplified) digests them, deconstructs them into pieces, and then spits …Because an LLM doesn't "learn" from training materials, it (vastly simplified) digests them, deconstructs them into pieces, and then s…
↗Not briefed
hn
hn-ai
1moClausal Prolog: A Neurosymbolic Prolog Built on Python—
↗Not briefed
bluesky
bluesky-ai
1moOpenAI sued for trademark infringement over 'Astra' AI model reut.rs/3Vuszv1OpenAI sued for trademark infringement over 'Astra' AI model reut.rs/3Vuszv1
↗Not briefed
bluesky
bluesky-ai
1moThis. There are people right now in the LLM society striving to develop models from scratch using only public domain data. Of course they'r…This. There are people right now in the LLM society striving to develop models from scratch using only public domain data. Of course t…
↗Not briefed
bluesky
bluesky-ai
1mo3D printers have now become zero-skill infrastructure when connected to an Ai. OpenSCAD + PrusaLink + Claude is a miracle in my lab. Highl…3D printers have now become zero-skill infrastructure when connected to an Ai. OpenSCAD + PrusaLink + Claude is a miracle in my lab.…
↗Not briefed