OpenAI, Anthropic and Google just pushed AI faster
AI just got a lot more agentic: OpenAI, Anthropic, and others are shipping tools that can use your computer, remember context, and run longer workflows. Here are the latest moves worth tracking today.[3][4][5]
🧵 1/5
OpenAI Codex overhaul: it can now use your computer, browse the web in-app, generate images, remember preferences, and run automations over days or weeks. That turns Codex into a full work partner, not just a coding helper.[3]
🧵 2/5
OpenAI also launched GPT-Rosalind, a life sciences reasoning model for biology, drug discovery, and translational medicine, with a Codex research plugin that connects to 50+ tools and data sources. Big signal for AI in science workflows.[3]
🧵 3/5
Anthropic’s Claude Opus 4.6 now has a 1M-token context window in beta, plus better coding, research, and long-task performance. It also adds adaptive thinking and context compaction, which should help agents stay on track longer.[4]
🧵 4/5
Claude Managed Agents are adding dreaming, outcomes, multiagent orchestration, and webhooks in beta. The bet is clear: agents that can learn from past sessions, check their own work, and split into specialists will move faster.[5] Which update stands out most to you? Reply or retweet with your take.[5]
🧵 5/5
Sign Up To Try Advanced Features
Get more accurate answers with Super Pandi, upload files, personalized discovery feed, save searches and contribute to the PandiPedia.