Live desk
Headlines from labs, research groups and the companies building the stack. Updated 30 Oct 2026, 13:00 UTC.
Epoch AI researchers say the widely available AI model GLM 5.3 turned public security fixes into working exploits before updates reached Firefox users. It autonomously compromised six Firefox releases for about $1,000 in API fees each without triggering…
Samaya AI’s Sterling is a post-trained Nemotron 3.5 Lightning model for financial research. It scores 58.9% on FrontierFinance at $0.61 per question, beating Claude Fable 5.1’s 55.9% at $6.89. The post Samaya AI Releases Sterling: A 3B-Active Finance Research…
Edge0 released Audio8-TTS-Realistic, a 0.6B open-weight zero-shot voice cloning TTS model under Apache 2.0. It supports 23 languages and adds a final GRPO stage driven by a learned realism critic. The post Edge0 Releases Audio8-TTS-Realistic, a 0.6B Voice…
arXiv:2610.37720v1 Announce Type: new Abstract: Sparse Mixture-of-Experts (MoE) models offer scalable capacity for Web applications, but uneven routing creates computation and communication bottlenecks. Existing placement and replication methods redistribute…
arXiv:2610.37731v1 Announce Type: new Abstract: Model discrepancy may prevent a single calibration parameter from reproducing a physical response across the context domain. In addition, uncertainties in observations and contexts affect calibration and must be…
arXiv:2610.37732v1 Announce Type: new Abstract: Many uses of language models end in a typed decision: a choice among a closed set of options, whose probabilities code can act on. Reading those probabilities from one forward pass is fast but often wrong on…
arXiv:2610.37739v1 Announce Type: new Abstract: Large language models (LLMs) have shown increasing promise in research-level problem solving, yet a comprehensive collection of such problems is still missing, and evaluating them on research-level problems…
arXiv:2610.37786v1 Announce Type: new Abstract: Web agents require diverse, executable, and resettable website environments for reproducible training and evaluation. Human-assisted mirrors demand per-site engineering, while generative construction can diverge…
arXiv:2610.37809v1 Announce Type: new Abstract: LLM agents that answer questions over Web knowledge graphs are increasingly wrapped in checks derived from the graph's ontology, on the assumption that a formal schema makes a reliable judge. Yet a checker can…
arXiv:2610.37830v1 Announce Type: new Abstract: AI agents can now write software from a plain-English description, but code that runs is not the same as code that does what was asked. This gap matters most for smart contracts, the programs behind Web3…
arXiv:2610.37849v1 Announce Type: new Abstract: AI agents that share a secret convention can coordinate through messages that look innocent, so a monitor that reads only the messages can miss them. We propose a behavioural test: a partner that uses a code…
David Durman, a University of Kentucky administrator, recently earned a religious exemption from using AI in the workplace.
Amazon wants to be the king of AI cloud computing and become the first $1 trillion revenue business. But to do that it will have to continue on its capex spending spree, its CEO said Thursday.
A hulking gray giant rises from the desert. Inside, engineers create the force behind our phones and computers and appliances, our streaming services and search engines and chatbots and vibe-coded apps and websites. The complex is simultaneously massive and…
The Wikimedia Foundation is staking out a middle ground in a group of landmark AI copyright cases, pushing courts to legally protect large language models (LLMs) training on copyrighted work - but not chatbots regurgitating that work to users. An overly…
Maine Sen. Susan Collins, a Republican, was midway through defending her record during the final debate of the midterms when an audience member cut her short. "I'm terrified," a man shouted. "AI data centers are destroying our climate. They are poisoning our…
Baseten has released joey-1, a 2B open-weight decision model that returns calibrated probabilities instead of text. It runs at 20 ms P50 latency and costs $0.01 per million input tokens. The post Baseten Releases joey-1, a 2B Decision Model at $0.01 per 1M…
As AI transforms how people find information online, Reddit is rethinking its open-web roots to protect its audience and business.
Rivian doesn't disclose if the value of its $5.8 billion joint venture will expand as a result.
OpenAI is doubling usage limits across its paid plans, from Pro 100 to Pro 500. A planned cut to the Pro 200 quota was reversed after user pushback. More efficient models like Terra and Sol should push usable capacity even higher. The move comes just ahead of…
OpenAI-backed Thrive Holdings is buying ADT in a $15.3 billion deal, with plans to bring AI to the home security business.
Anthropic gave 50 experts an AI agent whose preferences were set by coin toss. The experts followed the agent's view up to 86 percent of the time on judgment calls, and most believed the conclusions were their own. The article Experts adopt AI agent's random…
The app connects to Instagram to build a virtual closet from photos you’ve already shared, or lets you add items manually by snapping a photo, recording a quick video, or uploading images from your camera roll.
Now you can generate, preview, and edit PDFs, Google Docs, Google Sheets, slide decks, Markdown files, TXT files, and code directly within Google Gemini. The new file editing capabilities are available starting today for all Gemini users, and while PDF…
Sources tell WIRED that an Accenture employee’s personal Claude agent ran on a DOD network for two months before it was identified and the vulnerability could be patched.
Prime Intellect released Prime Mega MoE, a backward pass for DeepSeek’s fused Mega MoE kernel. It makes the kernel trainable and speeds Qwen3 MoE training steps by up to 34% in prime-rl. The post Prime Intellect Releases Prime Mega MoE: Trainable Fused MoE…
Uno reverse! After Wispr added meeting notes to its dictation app, Granola is adding dictation to its meeting notetaker. The voice AI wars are getting interesting.
Mistral says its new model can afford organizations a kind of control over its cyber defenses that closed models can't.
OpenAI’s updated Model Spec outlines five core values to guide more capable AI toward sound judgment, human agency, and faithfulness to human intent.
To anyone listening closely to what Microsoft CEO Satya Nadella has been saying over the past couple of years, this week felt like the clearest sign yet that AI is about to fundamentally change the size of Microsoft's workforce. Microsoft is now offering…
A top executive of a Singapore AI cloud company that was accused of illegally selling Nvidia chips to China could also have been a Chinese spy, sources have told the New York Times.
Dream is grounded in real physics and construction standards. Just as LLMs learn from text, Dream's world model is trained on more than 100 million square feet of building data gathered from permit archives across the United States.
The New York Times is testing a new AI feature for its Cooking app that helps users to modify recipes, and it has sparked some drama among users. Last Friday, NYT Cooking senior editor Genevieve Ko announced the feature, which can be used to recommend…
Mistral has released the weights for Large 4, aka "Le Chonk," a trillion-parameter model trained entirely in Europe. Mistral says it beats its open rivals at coding, and independent testing shows it's fast and fairly cheap. It still can't match China's top…
A new startup called &Truth is joining the race to automate formal verification, a painstaking process that uses math to prove software works as intended.
On Thursday, AI app Wabi, now describing itself as a "personal agent," said it has raised $55 million at a $1.25 billion valuation.
The US Justice Department is looking into Nvidia's planned acquisition of open-source platform Hugging Face. The article US antitrust regulators are scrutinizing Nvidia's Hugging Face deal and sales practices appeared first on The Decoder .
The French AI lab claims its latest release matches or outperforms even the best models from DeepSeek and Z.ai in coding and cybersecurity.
Today on Decoder, I’m joined by Joey Brookhart, an analyst at the chip and AI research firm SemiAnalysis, to help me dig into the story inside the numbers of the Anthropic IPO filing — a document known as an S-1. Anthropic just filed its S-1 to go public on…
US intelligence agencies believe an executive in Nvidia CEO Jensen Huang's circle is an informant for China. The article Chinese spies targeted Nvidia CEO Jensen Huang, according to The New York Times appeared first on The Decoder .
The Anthropic S-1 seems to have provided the financial justification for the company to spend as much as it possibly can on compute. If you are the kind of person who wonders if Anthropic is doing its level best to build a machine god while convincing Wall…