Thoughts on Figma AI’s Generative Interfaces

Avyay Kashyap
3 min readJul 1, 2024

--

The 26th of June, 2024, marked the beginning of a new wave of design. No longer would designers need to spend hours searching for the right component. Designers would now be able to upload an image and Figma would surface the components from your design system for you to design.

I’m kidding. No one is talking about this. Everyone’s talking about the new feature that allows designers to “Generate designs from text prompts”. There have been two types of reactions to this announcement — one of trepidation and concern, given the already saturated job market; others, however, show a lack of concern for the feature, recognising its limited impact on the workflows of experienced designers. And whether we like it or not, this day had to come. The moment ChatGPT captured people’s imagination, AI was going to be in every product.

But, as with politics and the economy, there is little benefit to taking things to the nth degree. A balance of healthy skepticism and optimism about embracing the oncoming shift in the multi-decade-long design process is necessary. I decided to go through Figma’s blog to learn more about their thoughts on how they approached designing these features.

Let’s discuss the “concerning” Elephant

Figma’s CTO, Kris Rasmussen, introduces the Generative Designs feature by bringing up the Blank Canvas Syndrome. It’s a challenge creatives encounter in figuring where to start, often because there are many complexities involved in the creative process, and fear of failure is real. There are tons of blogs on this topic, covering different strategies creatives have used to mitigate this problem. Figma suggests that, instead of being stuck, we should describe what we want, and the “feature will provide the first draft.”

How will this impact designers? Well, while I think this is great for getting that first idea out of the head, I believe there is a lot more nuance to the blank canvas syndrome, and merely generating a bunch of screens from the design library is not going to get designers anywhere close to getting rid of a creative block. It is darkest before dawn, and with the creative process, the anxiety that precedes the answer is often a necessary step in the process. The question I have at this point is “Can putting thought into the prompt be one outlet of expression that gets us a step closer to recognising the problem at hand?”

Will it eliminate the need for designers? I definitely do not think so, unless the organisation itself looks at designers as cogs who churn out serviceable artboards for developers to reference. Figma themselves suggest to look at this feature is as an inspiration generator— a better Dribbble, maybe. A quicker way to explore design directions. This still leaves the onus of decision making in the hands of designers — evaluators of work, selectively breeding for their context, and providing rationale for design choices. “Good artists copy and great artists steal” is a well known expression, and this makes copying a lot more accessible. So it very well might be the thing that lifts that floor of design.

Is the floor being lifted enough? For a large majority, especially those with resource constraints, acceptable design becoming commoditised will significantly improve the product’s value proposition. But that is a very low bar, which still leaves a lot of room for designers to add real value.

I understand the trepidation that surrounds this feature launch, but designers are more than just pixel pushers. The craft of design is valued for its grounding in cognitive theories and appreciation of contextual complexities. AI matches patterns, providing no rationale for design decisions. Designers figure what people want and how UI elements can support that.

What’s my take?

I tend to look at generative content as mere suggestions. Taking a leaf out of Hayes model of the writing process, one needs to evaluate content based on the motivations for generating the content, and keep evaluating to ensure it remains aligned with the larger goal at hand. As AI gets better at suggesting, and the craft of doing design begins to take a back seat, articulating the meaning behind the work and advocating for users with well founded arguments will come to fore. This just might be the development that saves design from itself.

--

--