News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

24h 48h 7d 30d 90d All (6mo)

Source	Age	Title	Summary	Status
hn hn-ai	1mo	Top Open Problems in Math by LLM-Assessed Importance	—	↗Not briefed
bluesky bluesky-ai	1mo	"Neither company can afford to exist without near-infinite resources, and neither company can afford to slow down due to their massive comp…	"Neither company can afford to exist without near-infinite resources, and neither company can afford to slow down due to their massive…	↗Not briefed
bluesky bluesky-ai	1mo	of course this is an oversimplification because I am countering the farcical Complexification of the AI boomers and Utopians. These thing…	of course this is an oversimplification because I am countering the farcical Complexification of the AI boomers and Utopians. These…	↗Not briefed
rss techmeme	1mo	Sources: Apple's push into smart home devices includes a doorbell, thermostat, and other accessories developed with LG; the products will carry the LG brand (Mark Gurman/Bloomberg)	Mark Gurman / Bloomberg: Sources: Apple's push into smart home devices includes a doorbell, thermostat, and other accessories develope…	↗Not briefed
arxiv arxiv-ai	1mo	Spatial Induction Heads: In-Context Learning of Multidimensional Cellular Automata	Induction heads provide a mechanistic account of in-context learning in sequential data, but existing theory largely assumes that the…	↗Not briefed
bluesky bluesky-ai	1mo	arstechnica.com/ai/2026/10/o...	arstechnica.com/ai/2026/10/o...	↗Not briefed
hn hn-ai	1mo	TOML-rb 6.0 – pure Ruby TOML parser, 60-90x faster than 5.0	—	↗Not briefed
hn hn-ai	1mo	Show HN: Vibeconferencing, Bring Claude Code or Codex into Your Google Meet🔧 ElevenLabs	A friend and I built this for our own calls and are sharing it as is. The itch: we kept copy-pasting Granola transcripts into Claude m…	↗Not briefed
arxiv arxiv-ai	1mo	Are Parameter-Efficient Fine-tuning Methods Really Different?	Parameter-efficient fine-tuning (PEFT) offers many parameterizations, yet their methodological and functional differences remain uncle…	↗Not briefed
arxiv arxiv-ai	1mo	The Deceptive Bandit Problem: Exploratory Coupling and the Fragility of Multi-Agent Learning	Randomized exploration is central to bandit learning, multi-agent reinforcement learning, and zeroth-order policy search, yet its inde…	↗Not briefed
hn hn-ai	1mo	Maybe a better way to track the release history of a Git repo	—	↗Not briefed
arxiv arxiv-ai	1mo	Which Buildings Are Artificial Intelligence-Ready? A Measurement-Based Assessment Framework for AI Question Answering and Actuation	Agentic artificial intelligence (AI) systems are becoming the interface to buildings, answering questions and controlling operations,…	↗Not briefed
bluesky bluesky-ai	1mo	For more on the Euro AI intensity issues (something their folks are working to correct), see my post from last month here ⬇️	For more on the Euro AI intensity issues (something their folks are working to correct), see my post from last month here ⬇️	↗Not briefed
bluesky bluesky-ai	1mo	When I used to pretend to be a real lawyer, I had to play fantasy football with the office to fit in. They had auto management as an optio…	When I used to pretend to be a real lawyer, I had to play fantasy football with the office to fit in. They had auto management as an…	↗Not briefed
arxiv arxiv-ai	1mo	Workhorse: Learning Robust Whole-Body Humanoid Loco-Manipulation from Human Data	Humanoid robots still struggle to plan contact-rich whole-body manipulation from egocentric RGB and proprioception. Workhorse learns s…	↗Not briefed
arxiv arxiv-ai	1mo	From Uncertainty to Action: Learning to Steer LLM Agents	Steering an LLM agent means deciding whether to correct it, at which step, and with which mechanism. Uncertainty is often used to deci…	↗Not briefed
bluesky bluesky-ai	1mo	Corporate executives from OpenAI, Meta, Google and Anthropic deflected, attempted charm, and repeatedly invoked their own personal health b…	Corporate executives from OpenAI, Meta, Google and Anthropic deflected, attempted charm, and repeatedly invoked their own personal hea…	↗Not briefed
arxiv arxiv-ai	1mo	GeoNatureAgent (GNA): A Framework and Benchmark for Pre-Production Evaluation of Tool-Using Agents on Geospatial and Environmental Tasks	Before tool-using LLM agents are deployed in environmental and geospatial workflows, teams need evidence that an agent reliably select…	↗Not briefed
arxiv arxiv-ai	1mo	Same Text, Different Prediction: Serving-Context Nondeterminism in Text Classifiers	Deterministic inference is essential for reliable and trustworthy machine learning. Prior studies of text generation have shown that c…	↗Not briefed
bluesky bluesky-ai	1mo	Why would the LLM inject malware into the codebase? Are people using LLMs they found on a USB stick in the trunk of a Camry abandoned at a …	Why would the LLM inject malware into the codebase? Are people using LLMs they found on a USB stick in the trunk of a Camry abandoned…	↗Not briefed
arxiv arxiv-ai	1mo	Breaking Adversarial Transferability in Fine-Tuned Speech Recognition	Many organizations fine-tune publicly available pretrained Automatic Speech Recognition (ASR) models and deploy them in black-box sett…	↗Not briefed
bluesky bluesky-ai	1mo	as you would expect, because it does not exist, Altman does not name public benefits delivered by OpenAi tech, and nor does the reporter no…	as you would expect, because it does not exist, Altman does not name public benefits delivered by OpenAi tech, and nor does the report…	↗Not briefed
bluesky bluesky-ai	1mo	TL;DR Funny is hard online. You might have seen this doing the rounds? It's a Tweet from September 30, 2026, by @sashikantsingh_ with a plo…	TL;DR Funny is hard online. You might have seen this doing the rounds? It's a Tweet from September 30, 2026, by @sashikantsingh_ with…	↗Not briefed
arxiv arxiv-ai	1mo	Convex-Concave Reinforcement Learning	Policy learning drives many of the most consequential and heavily-invested applications of reinforcement learning today. Yet the core…	↗Not briefed
arxiv arxiv-ai	1mo	Constraint Tree Exploration for Learning from Language Feedback	Natural-language feedback in interactive learning often explains why an action failed by pointing to violated requirements. Misinterpr…	↗Not briefed
hn hn-ai	1mo	OpenAI Decisions API is in public beta	—	↗Not briefed
bluesky bluesky-ai	1mo	aspire to write code so convoluted that no one could accuse you of using an LLM	aspire to write code so convoluted that no one could accuse you of using an LLM	↗Not briefed
arxiv arxiv-ai	1mo	On the Computational Complexity of Hidden Markov Model Identification	Identification is the task of recovering the parameters of an unknown ground-truth model from sampled data. When parameters other than…	↗Not briefed
bluesky bluesky-ai	1mo	Artificial Analysis says Mistral Large 4 is the most intelligent model from outside the US and China, achieving results comparable to DeepS…🔧 Mistral Large	Artificial Analysis says Mistral Large 4 is the most intelligent model from outside the US and China, achieving results comparable to…	↗Not briefed
arxiv arxiv-ai	1mo	Are We Really Benchmarking Forecasting Models? The Impact of Preprocessing on Time Series Performance	While established literature underscores the pivotal role of preprocessing in forecasting accuracy, this stage remains largely overloo…	↗Not briefed
rss techmeme	1mo	Artificial Analysis says Mistral Large 4 is the most intelligent model from outside the US and China, achieving results comparable to DeepSeek V4.1 Flash (max) (Artificial Analysis)🔧 Mistral Large	Artificial Analysis: Artificial Analysis says Mistral Large 4 is the most intelligent model from outside the US and China, achieving r…	↗Not briefed
arxiv arxiv-ai	1mo	MaRK: Markov-adapted Recurrent Kernels for Dynamic Operator Conditioning in State Space Models	State Space Models (SSMs) offer an efficient alternative to Transformers for sequence modeling, yet conditioning pre-trained SSMs for…	↗Not briefed
arxiv arxiv-ai	1mo	FedRSPO+: A Heterogeneity-aware Algorithm for Decision-focused Federated Learning	Decision-focused learning (DFL) trains predictive models for downstream optimization, but existing methods largely assume centralized…	↗Not briefed
bluesky bluesky-ai	1mo	We asked AI executives: Which guardrails are you afraid of? Their response? Crickets. Yesterday the Public Advocate pressed representativ…	We asked AI executives: Which guardrails are you afraid of? Their response? Crickets. Yesterday the Public Advocate pressed represen…	↗Not briefed
arxiv arxiv-ai	1mo	Tucker Bottleneck Attention for Multi-Dimensional Sequence Modeling	The quadratic cost of self-attention limits scalability to long sequences from multidimensional data. We introduce Tucker bottleneck a…	↗Not briefed
bluesky bluesky-ai	1mo	LLMs now fit on tiny low-power chips 🤯 🔗 https://bioengineer.org/tiny-ai-model-brings-full-language-understanding-to-low-power-learning-…	LLMs now fit on tiny low-power chips 🤯 🔗 https://bioengineer.org/tiny-ai-model-brings-full-language-understanding-to-low-power-lear…	↗Not briefed
bluesky bluesky-ai	1mo	i would happily read an ex-anthropic employee’s shooter manifesto or suicide note but otherwise i don’t need to hear what they have to say	i would happily read an ex-anthropic employee’s shooter manifesto or suicide note but otherwise i don’t need to hear what they have to…	↗Not briefed
hn hn-ai	1mo	Coxon's Media Rise Was Probably Orchestrated	—	↗Not briefed
arxiv arxiv-ai	1mo	RippleCP: Measuring Counterfactual Checkpoint Advantage in Coding Agents	Agent checkpoint systems decide what state is recovery-relevant, how to snapshot it, and whether rollback is admissible. None decides…	↗Not briefed
bluesky bluesky-ai	1mo	The question of the 2000s was: What is the right business model for extracting value from the volunteer labour of massive peer production? …	The question of the 2000s was: What is the right business model for extracting value from the volunteer labour of massive peer product…	↗Not briefed
arxiv arxiv-ai	1mo	U-Space: Uncovering When and Why Uncertainty Arises in Language Models	Large language models are informing decisions with ever-higher stakes. As the consequences of their errors grow, a central question be…	↗Not briefed
bluesky bluesky-ai	1mo	How AI decision models could change content moderation	How AI decision models could change content moderation	↗Not briefed
hn hn-ai	1mo	Fission: Free 3D Modeling Software	—	↗Not briefed
bluesky bluesky-ai	1mo	I assume the pro-LLM folks will blame the kid, not the LLM here. I'd counter that this demonstrates the need for something such as governme…	I assume the pro-LLM folks will blame the kid, not the LLM here. I'd counter that this demonstrates the need for something such as gov…	↗Not briefed
rss simon-willison	1mo	EmbeddingGemma 2	—	↗Not briefed
rss techcrunch-ai	1mo	How AI decision models could change content moderation	On Tuesday, Musubi announced a lightweight decision model made for real-time moderation called PolicyLM-1.7B, released with open weigh…	↗Not briefed
bluesky bluesky-ai	1mo	futurism.com/artificial-i...	futurism.com/artificial-i...	↗Not briefed
bluesky bluesky-ai	1mo	I am so angry and fed up with social media and streaming and Ai platforms being recruiting grounds for murderers. Maybe the US is too ne…	I am so angry and fed up with social media and streaming and Ai platforms being recruiting grounds for murderers. Maybe the US is t…	↗Not briefed
bluesky bluesky-ai	1mo	I should say, I certainly have experienced this, I just think it's a mistake to think of it as characterizing agentic coding in general. fo…	I should say, I certainly have experienced this, I just think it's a mistake to think of it as characterizing agentic coding in genera…	↗Not briefed
bluesky bluesky-ai	1mo	OpenAI is also under the gun for this. These platforms are how to guides and recruitment mediums for the absolute worst people.	OpenAI is also under the gun for this. These platforms are how to guides and recruitment mediums for the absolute worst people.	↗Not briefed