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
2moDiscovering Interpretable Algorithms by Decompiling Transformers to RASP [video]—
↗Not briefed
bluesky
bluesky-ai
2moThe judge has also allowed the parties to update the website with the statement that initial payments (for the first $1.05 billion put in b…The judge has also allowed the parties to update the website with the statement that initial payments (for the first $1.05 billion put…
↗Not briefed
rss
techcrunch-ai
2moThe AI data center boom is colliding with cities scarred by big industryNational outcry against data center construction has spread to Philadelphia, where officials suggested possible construction in a neig…
↗Not briefed
bluesky
bluesky-ai
2moa little chuckle for anyone who thinks anthropic has the best model and the worst harnessa little chuckle for anyone who thinks anthropic has the best model and the worst harness
↗Not briefed
hn
hn-ai
2moBranch Divergence in SIMT Machines: From CFG to Compiler Divergence Analysis [pdf]—
↗Not briefed
bluesky
bluesky-ai
2moIn the Anthropic settlement, Judge Araceli Martinez-Olguin has agreed to the parties' request to update the website, which means that the j…In the Anthropic settlement, Judge Araceli Martinez-Olguin has agreed to the parties' request to update the website, which means that…
↗Not briefed
arxiv
arxiv-ai
2moUncertainty-Aware Continual Learning for Open-World Intent Discovery Under an evolving Label SpaceReal-world intelligent systems increasingly operate under open-world conditions, where user intents are not fixed or exhaustively know…
↗Not briefed
arxiv
arxiv-ai
2moDo Frontier Models Seek Safety Evidence Before Acting?🔧 GPT-5Frontier models are often evaluated on how they respond to safety information once it is already in context. We study an earlier decis…
↗Not briefed
arxiv
arxiv-ai
2moThe Inference Engineering Pareto Atlas: Which Optimizations Dominate the Cost, Quality, and Latency Frontier?🔧 vLLMLLM inference optimizations report speedups on different models, GPUs, prompts, and quality metrics, making them hard to compare or co…
↗Not briefed
bluesky
bluesky-ai
2moYes, the American AI industry is massively intertwined with the people who brought you Aella’s birthday gangbang. Because they are a mille…Yes, the American AI industry is massively intertwined with the people who brought you Aella’s birthday gangbang. Because they are a…
↗Not briefed
bluesky
bluesky-ai
2moChatGPTとGPTは「サービス」と「AIモデル」の違い。 ClaudeとClaude Codeは「AIアシスタント」と「そのAIを使って作業を進めるためのツール」の違い。 ( ˙꒳​˙ )フムフム…🔧 ChatGPTChatGPTとGPTは「サービス」と「AIモデル」の違い。 ClaudeとClaude Codeは「AIアシスタント」と「そのAIを使って作業を進めるためのツール」の違い。 ( ˙꒳​˙ )フムフム…
↗Not briefed
bluesky
bluesky-ai
2moYes, the American AI industry is massively intertwined with the people who brought you Aella’s birthday gangbang. Because they are a mille…Yes, the American AI industry is massively intertwined with the people who brought you Aella’s birthday gangbang. Because they are a…
↗Not briefed
bluesky
bluesky-ai
2mofolks we've got a completely unironic "why aren't AI companies using the three laws of robotics from the classic series Absolutely Do Not U…folks we've got a completely unironic "why aren't AI companies using the three laws of robotics from the classic series Absolutely Do…
↗Not briefed
bluesky
bluesky-ai
2moNew: when a hacking blog says "no politics" (👀), my Anthropic update, more OpenAI hack attacks emerge as Altman peddles OpenAI’s cyber-def…New: when a hacking blog says "no politics" (👀), my Anthropic update, more OpenAI hack attacks emerge as Altman peddles OpenAI’s cybe…
↗Not briefed
rss
marktechpost
2moInside NVIDIA’s cuDNN Graph API: Fusion, Autotuning, and Plan Reuse with cuDNN FrontendLearn how to leverage NVIDIA’s cuDNN Frontend Graph API to build custom kernel fusions, autotuning engine configurations, FP8-style ep…
↗Not briefed
hn
hn-ai
2moI asked Claude to build AI generator on my own hardware—
↗Not briefed
rss
techmeme
2moOpenRouter users spent more on OpenAI's models than on Anthropic's in the week of September 7, the first time that happened since the week of February 26, 2024 (@openrouter)🔧 OpenRouter@openrouter: OpenRouter users spent more on OpenAI's models than on Anthropic's in the week of September 7, the first time that happen…
↗Not briefed
bluesky
bluesky-ai
2moPocock says Labor easing copyright protections for AI datacentre investment would throw creatives ‘under the bus’Pocock says Labor easing copyright protections for AI datacentre investment would throw creatives ‘under the bus’
↗Not briefed
bluesky
bluesky-ai
2momy VMware "ai safety" product was - gateway for all LLM calls - webhook for input screen - webhook for output screen - log (securely!) by …my VMware "ai safety" product was - gateway for all LLM calls - webhook for input screen - webhook for output screen - log (securely!…
↗Not briefed
arxiv
arxiv-ai
2moWho Judges Matters: Measuring Family-Conditioned Preference in LLM-as-Judge Panels🔧 ConsensusWho the judge is can affect an LLM-as-judge result, but measuring that effect without confusing it with candidate quality is difficult…
↗Not briefed
bluesky
bluesky-ai
2moI can imagine being amused by this a few years ago - of toying with the robot. But every interaction with AI burns carbon and water, and he…I can imagine being amused by this a few years ago - of toying with the robot. But every interaction with AI burns carbon and water, a…
↗Not briefed
bluesky
bluesky-ai
2moAnd the balance can shift; the LLM itself could be reduced to a HCI interface, while most of the cognition happens on the backend.And the balance can shift; the LLM itself could be reduced to a HCI interface, while most of the cognition happens on the backend.
↗Not briefed
bluesky
bluesky-ai
2moi am excited by the potential of reinforcement learning for calibrated decisions! signed up for the waitlist, answered some questions. the…i am excited by the potential of reinforcement learning for calibrated decisions! signed up for the waitlist, answered some questions…
↗Not briefed
hn
hn-ai
2moShow HN: HonestSky, a free iOS weather app without ads or subscriptionsHi HN, I’m Sean. I built HonestSky because I was tired of weather apps that were slow to load, covered in ads some to the point of bei…
↗Not briefed
bluesky
bluesky-ai
2moYes. 😁 This inverts the pattern; stick this into an LLM harness and you give it something like embodied cognition.Yes. 😁 This inverts the pattern; stick this into an LLM harness and you give it something like embodied cognition.
↗Not briefed
arxiv
arxiv-ai
2moSNOMED CT Concept Recommendation from Masked Clinical ContextStandardizing clinical language to SNOMED CT supports interoperability, analytics, and reusable phenotyping, but concept recommendatio…
↗Not briefed
hn
hn-ai
2moFor those into GenAI entertainment, I am excited about this hackathonGossip Goblin hackathon taking place at Google DeepMind in SF on 10/20 -- goal is for founders building in the future of genAI en…
↗Not briefed
hn
hn-ai
2moClaude Roulette – chat with strangers while you slop—
↗Not briefed
arxiv
arxiv-ai
2moAfriSyCo: Measuring Assertive Framing, Verification, and Wording Sensitivity Around African-Language ContentAfriSyCo studies answer switching around African-language factual content with two complementary layers: native-language follow-ups an…
↗Not briefed
bluesky
bluesky-ai
2moI don’t think generative AI will wipe out all life on earth but I do think it has the capacity to cause a lot of arbitrary chaos because ve…I don’t think generative AI will wipe out all life on earth but I do think it has the capacity to cause a lot of arbitrary chaos becau…
↗Not briefed
hn
hn-ai
2moLocal LLM engine where a tool call cut by the token limit still parses—
↗Not briefed
bluesky
bluesky-ai
2moreading stuff by AI true believers makes me feel insane. i read a post about a "fable level model" that "only makes decisions" (no text out…reading stuff by AI true believers makes me feel insane. i read a post about a "fable level model" that "only makes decisions" (no tex…
↗Not briefed
hn
hn-ai
2moAqueduct Water Risk Atlas—
↗Not briefed
bluesky
bluesky-ai
2moThe statement from former Anthropic researcher Jacob Coxon that AI “could kill us by the end of the decade” has triggered the most hostile …The statement from former Anthropic researcher Jacob Coxon that AI “could kill us by the end of the decade” has triggered the most hos…
↗Not briefed
bluesky
bluesky-ai
2moI find the claim that an LLM "can't" hallucinate eyebrow-raising. How is that possible?I find the claim that an LLM "can't" hallucinate eyebrow-raising. How is that possible?
↗Not briefed
bluesky
bluesky-ai
2moAlso this AI curtailment narrative is so stupid. Anthropic have signed several multiple $100bn deals for compute from now to 2029. Give me …Also this AI curtailment narrative is so stupid. Anthropic have signed several multiple $100bn deals for compute from now to 2029. Giv…
↗Not briefed
bluesky
bluesky-ai
2moYeah I run PR at anthropic and I knew we had to get universally loved figure Will Stancil as a secret spokesmanYeah I run PR at anthropic and I knew we had to get universally loved figure Will Stancil as a secret spokesman
↗Not briefed
bluesky
bluesky-ai
2moJag tror din LLM hallucinerade upp det där receptet när du frågade "snälla söta llm sir madam jag är lite hungrig och är ganska sugen på nå…Jag tror din LLM hallucinerade upp det där receptet när du frågade "snälla söta llm sir madam jag är lite hungrig och är ganska sugen…
↗Not briefed
arxiv
arxiv-ai
2moSFT or RL for Tool-Calling Agents? A Controlled Study Across Data, Method, and ScaleLimited controlled evidence exists on how training data, adaptation method, and model scale jointly affect tool-calling performance in…
↗Not briefed
bluesky
bluesky-ai
2moClaudeにコード書かせてて、ふと思い出して自分のgithub見に行ったら要らない大昔のrepoばっかり。全部消去した。 ソフトウェア業リタイアしてからろくに見なかったし、もう他人とプログラミングで関わる事もない。 そもそもこれだけAIがコード書ける時代に、もう自分でプログ…Claudeにコード書かせてて、ふと思い出して自分のgithub見に行ったら要らない大昔のrepoばっかり。全部消去した。 ソフトウェア業リタイアしてからろくに見なかったし、もう他人とプログラミングで関わる事もない。 そもそもこれだけAIがコード書ける時代に、もう自…
↗Not briefed
bluesky
bluesky-ai
2mo“AI labs should not fall into a ‘false choice’ between the speed of innovation and the ability to develop safe products” -Jensen Huang, Sep…“AI labs should not fall into a ‘false choice’ between the speed of innovation and the ability to develop safe products” -Jensen Huang…
↗Not briefed
bluesky
bluesky-ai
2molike training an LLM on data generated by an LLM?like training an LLM on data generated by an LLM?
↗Not briefed
hn
hn-ai
2moGerman Rheinmetall open-sources its Battlesuite connected weapon system protcol—
↗Not briefed
arxiv
arxiv-ai
2moLearning Heterogeneous Preferences🔧 ModalLearning from human feedback has become a central paradigm for training modern AI systems, where models of human utility are used as r…
↗Not briefed
bluesky
bluesky-ai
2moNvidia CEO says battle over AI innovation and safety is ‘false choice’ ft.trib.al/bhOrdVhNvidia CEO says battle over AI innovation and safety is ‘false choice’ ft.trib.al/bhOrdVh
↗Not briefed
bluesky
bluesky-ai
2moWhen the department of war is declaring Effective Altruism a communist conspiracy that must be crushed, Yud is chastising the Anthropic CEO…When the department of war is declaring Effective Altruism a communist conspiracy that must be crushed, Yud is chastising the Anthropi…
↗Not briefed
hn
hn-ai
2moIntel reportedly cans 12Xe option for Nova Lake-S desktop—
↗Not briefed
arxiv
arxiv-ai
2moPrimeScientist: Strategic Allocation of Research Effort in Autonomous ResearchAutonomous research agents aim to automate scientific workflows, from proposing ideas to conducting experiments and analyzing results.…
↗Not briefed
arxiv
arxiv-ai
2moSharp margin-based generalization bounds for realizable SVMLet the exact homogeneous hard-margin support vector machine be trained on \(m\) independent observations from a Borel probability law…
↗Not briefed
arxiv
arxiv-ai
2moRoboVAD: A Large Cross-Domain Evaluation Benchmark for Anomaly Detection in Robotic Arm Manipulation VideosVideo anomaly detection (VAD) is an actively studied task, having wide applications in typical scenarios such as public surveillance a…
↗Not briefed