This project is suspended: no new articles or editions will be published. The archive stays available.
74 articles
At an internal town hall, Meta CEO Mark Zuckerberg conceded that AI agent development has not "accelerated in the way" the company expected, despite a $145 billion infrastructure commitment and a major workforce reorganization.
Read more →Anthropic released Claude Sonnet 5 on June 30, delivering near-Opus-level reasoning and agentic capabilities at introductory API pricing of $2 per million input tokens.
Read more →OpenAI has released GeneBench-Pro, a 129-problem benchmark that tests whether AI agents can exercise the judgment-intensive analytical reasoning required for real-world computational biology research.
Read more →Alibaba's Qwen team released a language world model that learns to predict agent environment states across seven domains, offering a new approach to training AI agents without live systems.
Read more →At its annual Config conference on June 24, Figma introduced Motion, a built-in timeline animation tool, and an upgraded Design Agent with reusable skills, external connectors to Notion and GitHub, and AI-generated shader fills.
Read more →Liquid AI released LFM2.5-230M, a 230-million-parameter open-weight model that runs AI agent tasks at 213 tokens per second on a phone CPU and 42 tokens per second on a Raspberry Pi — without a cloud connection.
Read more →ByteDance launched Doubao Seed 2.1 Pro and Turbo on June 23, claiming top scores on coding and agent benchmarks and parity with GPT-5.5 and Claude Opus 4.7.
Read more →An AI agent is software that can plan, make decisions, and take actions on your behalf — going far beyond a basic chatbot. Here is how they work and where to try one.
Read more →A multi-institution index of 30 deployed AI agents finds that 25 lack published internal safety results and 23 have no third-party testing, raising urgent questions as agentic systems take on increasingly consequential tasks.
Read more →Amazon's new Alexa+ Agentic Ads format lets customers move from seeing an ad to completing a purchase entirely within a conversation — no app-switching required.
Read more →Alibaba's Qwen team has released Qwen-AgentWorld, a family of open-source language world models that predict how digital environments respond to AI agent actions, with its 397B variant outscoring GPT-5.4 on the new AgentWorldBench benchmark.
Read more →