Simon Willison used Claude Opus 5.5 in Claude Code to build Chirp Pong, a static web game in which two phones exchange game state through audio chirps picked up by the microphone. He reports that the experiment works.
Meta's engineering team explains how it built Forum, a new app for deeper discussion in Facebook groups, by bringing an AI agent into the development cycle. The post shares lessons from that change in how the team worked.
The new Codex release lets users send a follow-up while the agent is still responding to change its course, and adds a searchable /config settings screen. It also runs independent read-only tools in parallel and adds a configurable leader key.
IBM argues that as AI agents take on more work, companies need an operating model linking agent actions to customer outcomes, infrastructure and financial accountability. Its approach folds automation, data and operational signals into one framework.
With the ChatGPT Atlas browser gone, OpenAI is upgrading ChatGPT's Chrome extension so the assistant can see the tabs a user already has open when Chrome is tagged in a conversation. It works from ChatGPT's web and desktop apps.
Anthropic has taken Code Review, its automated pull-request reviewer for Claude Code, out of research preview for Team and Enterprise customers, offering eligible enterprises up to $50,000 in free credits. Companies using GitHub Enterprise Cloud with data residency, a setup common among European firms, are excluded from the offer.