The Future of Design Tools: From Figma's Complexity to Claude's Simplicity

Yesterday, I experimented with Claude Design and developed a theory about the future of design tools. As design teams grew within engineering organizations, the need for systematization became paramount, leading to the creation of Figma's unique primitives like components and styles. However, this complexity has resulted in a convoluted system that requires specialized roles just to manage it. Figma's victory over Sketch by becoming the 'source of truth' came at a cost, as its closed and undocumented format excluded it from the training data of modern AI models, which are more aligned with code.

At work, we've often had to back-port design changes from code to Figma, a tedious process that highlights the inefficiencies of Figma's system. Even Figma's own design system, with its numerous variables and complex structures, exemplifies the challenges of debugging and managing designs. As the industry shifts back to code as the primary source of truth, Figma's manual system seems outdated.

I foresee design tools diverging into two paths: one that embraces code, like Claude Design, and another that focuses on pure creative exploration without code constraints. Claude Design aligns with the 'truth to materials' principle, being transparent about its HTML and JS foundation. Its integration with Claude Code offers a seamless feedback loop between design and implementation, reducing friction.

The second path will likely be a tool that allows for unrestricted creative exploration, potentially as an iPad app or a high-fidelity compositing tool like Photoshop. This tool would liberate designers from the limitations of CSS effects, allowing for more imaginative creations.

Figma's moment of reckoning is near, akin to Sketch's past challenges. To Figma, I say this situation could have been avoided if you had hired me. To Sketch, it's time to innovate and add exciting new features. And to my mom, sorry for the language.