Video · Riley Brown ·
Claude Just Changed EVERYTHING (Here’s What’s New)
In short
The video surveys new AI-agent capabilities across Claude, Gemini, GrokBot, and Dots, with emphasis on Claude’s workspace artifacts, meeting participation, model tradeoffs, and computer control. It argues that agents are becoming operational coworkers, but payment restrictions and platforms blocking agents remain major barriers to running businesses with them.
Key points
- Claude added Sheets, Whiteboards, and Data artifacts, plus Claude Meetings that can create and update artifacts during calls.
- Fable 5.2 is positioned as a higher-quality but twice-as-expensive option than Opus 5.5; Mythos 5.2 is restricted to vetted organizations for cybersecurity work.
- Gemini 4 Carbon is described as Google’s strongest model but not better than Opus 5.5 in the creator’s tests, despite temporarily costing half as much.
- GrokBot gained access to Mac apps, Slack improvements, and Tailscale support for private MCP servers and internal tools.
- Dots can coordinate multiple computers, use ChatGPT Wallet for approved purchases in the US, and run faster with GPT 6.1 Tera.
- United’s restriction of AI agents booking flights illustrates a broader conflict between agent convenience and companies’ pricing and advertising models.
- The recommended reading examines forensic evidence that Chinese AI labs may have trained models on export-restricted Nvidia Blackwell chips.
Written by an AI model from the video’s captions and description.
Claude is becoming an entire workspace. In this week's AI news update, I walk through Claude Sheets, Whiteboards, Data, and Meetings, plus Fable 5.2 and Mythos 5.2. I break down what these updates mean for using agents at work and when I use Fable instead of Opus.
In this story
A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.


