Anthropic has upgraded its Claude assistant with a new ability to see—and show—what you’re talking about. The AI can now generate custom charts, diagrams, and other visualizations directly within a chat window, moving beyond text to illustrate concepts on the fly.

If the context of a discussion suggests a picture would help, Claude will insert a relevant visual automatically. A conversation about the periodic table, for instance, might prompt the AI to produce an interactive version you can click for details. Users can also request a specific chart or diagram directly. These in-line visuals are ephemeral, changing or vanishing as the dialogue evolves, a contrast to the more persistent creations in Claude’s separate "artifacts" panel.

The update, now rolling out to all users by default, arrives as AI assistants increasingly embrace multimodality. Just this week, OpenAI demonstrated ChatGPT creating interactive science visuals, while Google’s Gemini has similar educational capabilities. Anthropic’s integration is distinct for weaving these graphics seamlessly into the conversational flow, allowing for real-time adjustments upon request. It’s a step toward AI that doesn’t just answer questions, but actively illustrates its reasoning.

Source: The Verge