Block's Dessa Speech Model Learned to Talk From Sound Alone: What It Does and Who Can Use It
Block's Dessa speech model learned to talk from sound alone, with no text. Here is what the research preview does, who built it and who can use it now.
Articles
Kimi K3.1 Pro Launches With a 1M-Token Context and Open Weights Due 4 November: Benchmarks, Access and Caveats
Kimi K3.1 Pro scores 65.7% on Terminal-Bench 4.0, beating GPT-6 Astra on Moonshot's numbers. Here is how to use it now and what the open weights mean.
Agentic AI Foundation Adds Agent Skills, Open Responses, OpenEnv and OpenSandbox: What Each One Does for Builders
The Agentic AI Foundation takes in four projects from Anthropic, OpenAI, Hugging Face and Alibaba. Here is what each does and why it matters if you build agents.
Microsoft AI Diagnosis Study: MAI-DxO Beat Doctors 83% to 36% on Hard Cases. What It Does and Doesn't Prove
Microsoft AI diagnosis tool MAI-DxO solved 83% of hard cases versus 36% for doctors working alone. Here is how it works, the catches, and what comes next.
OpenAI Decisions API Explained: Fast Yes-or-No AI Calls at $0.10 per Million Tokens
The OpenAI Decisions API is now open to all developers. It answers bounded questions up to 10x faster than GPT-6 Luna and charges nothing for output tokens.
Canva Craft Explained: What Canva's First Image Model Does, Where to Find It and What It Means for Designers
Canva Craft is Canva's first in-house image model: 2K images, 30 aspect ratios, readable text and layered edits, free in Dream Lab and via the Canva API.
1Password for Agents Explained: How AI Assistants Can Log In Without Ever Seeing Your Password
1Password for Agents lets Claude Duet, ChatGPT Work, Muse and Grok Bot sign in for you after you approve, with the password never shown to the AI agent.
Uber ChatGPT Work Rollout: All 30,000 Staff Get OpenAI's Agent After a 1,000-Person Pilot
Uber ChatGPT Work rollout: all 30,000 Uber staff now get OpenAI's workplace agent after a 1,000-person pilot in which 80% of early users were not engineers.
Nirva AI Pendant: The $249 Wearable That Only Listens to Its Owner, and What It Does With Your Voice
The Nirva AI pendant records only its wearer's voice and turns it into insights. Price, battery, privacy defaults and the $33m behind it, explained.
AI Shopping Agents Get a Front Door at Walmart While Amazon Keeps Them Out
Walmart's Agent Gateway lets approved AI shopping agents search and buy for you, with spending limits and separate profiles, while Amazon keeps agents blocked.
OnStar AI Assistant: What GM's New In-Car AI Does, When It Arrives and What It Means for Privacy
GM is turning OnStar into an AI assistant. Here is what the OnStar AI Assistant does, when it reaches cars and phones, and what GM says about your privacy.
Recursive Flint: The Free Coding Model That Runs on a Laptop, and Was Mostly Built by AI
Recursive Flint is a free, Apache 2.0 coding model that runs offline on a laptop. What it does, how fast it is, its limits, and what it means for developers.
Claude Duet Explained: Anthropic's Background AI Agent, What It Does and Who Can Use It
Claude Duet is Anthropic's new background AI agent that watches your apps and does tasks while you're away. What it does, its limits, and who can get it now.
OpenAI Models on Google Cloud: What Changes for Developers and Businesses
OpenAI models on Google Cloud arrive in Model Garden within weeks, with Codex and existing spend commitments. Here is what builders and IT buyers should know.
OpenAI Buys Orchestra Bio to Put AI Agents Inside Drug Development
OpenAI has bought Orchestra Bio, whose AI agents manage drug development programmes, and is folding the team into GPT-Rosalind and Rosalind Workbench.