🤿 Dive Club | Will Figma become an awkward middle ground?

🌈 Abstract

The article discusses a trend observed among designers who can code, where they spend more time sketching and less time in design tools like Figma. It explores the potential impact of AI on the design process, particularly in the context of product design, and proposes a different approach to leveraging AI in design workflows.

🙋 Q&A

[01] The Trend of Designers Who Can Code

1. What is the trend observed among designers who can code?

• Designers who can code spend more time sketching and less time in design tools like Figma.
• They can make a scribble and skip the Figma stage to jump directly into code.
• They quickly upgrade to interactive prototypes with pure CSS and HTML, which is more like sketching with code.

2. How does this trend compare to the author's typical design process?

• The author already follows this process for designing websites, where they sketch out concepts and then build them directly in Framer.
• However, the author acknowledges that this approach may fall outside of the core skillset for product design.

[02] The Impact of AI on Design Workflows

1. How does the author view the current state of Figma and design tools?

• The author believes that the current state of Figma feels like an "awkward middle ground" and that AI will change this.
• The reason Figma has over 4 million users is that it takes most people too long to code their designs, so they compromise by drawing "pretty pictures" in Figma.

2. What are the author's thoughts on the current AI tools for design?

• The author believes that most new AI tools live in the top left quadrant of the fidelity spectrum, which feels more like a "toy" than something that will meaningfully impact the design process.
• The author sees the current AI tools as "half-baked technology in search of a problem" and functionality geared toward non-designers.

3. What is the author's proposed approach to leveraging AI in design workflows?

• The author is more intrigued by a different way to leverage AI, where the focus is on generating usable code from a defined checkpoint in the design process.
• The author suggests a tool that sits somewhere between Balsamiq and tldraw, which would provide more structure for AI models than natural language and create a defined checkpoint in the design process where UX is the core deliverable.
• This approach would isolate critical thinking from the outputs that AI can excel at, such as turning wireframes into frontend code, wielding and extrapolating design systems, and creating beautiful visuals from screenshots and mood boards.