The State of AI in UI/UX Design
Advances in AI give UX designers an automated sidekick with a number of tangible benefits.
I’ve spent countless hours red-lining specs and working to communicate information from those specs to the development environment. It’s tedious, time-consuming, and arguably not the most creative use of my or my team’s time. But it’s necessary — as is adapting, translating, and rationalizing that work for localization, legibility, and accessibility once the designs are published.
Now, though, these workflows are getting the artificial intelligence (AI) and machine learning treatment, freeing designers like me to do more creative and customization work — and letting the machine handle those never-ending translations.
Though still a relatively new sidekick, AI and machine learning advances tied to UI/UX design are already exhibiting a number of tangible benefits. Here are just a few:
1. Reducing mundane, repetitive tasks
The red-lining example is just one of many repetitive, mundane tasks UI/UX designers know all too well. Already, though, AI and machine learning are having a profound effect on how much time we spend working through these processes.
More and more, designers are leveraging AI and machine learning to generate components either in code from a design document or in a design document from code. In the past, we were stuck doing manual translations, which could be very time consuming. This technology is a huge win and something more and more UX designers are happily turning over to AI and machine learning systems.
Airbnb is a great example. Their AI-driven UI technology identifies design sketches and converts them to code in real-time. There’s tremendous potential — the same technology could easily be used to live-code prototypes based on sketches and whiteboard drawings. A mock-up could become component specs instantly.
On my end of things, that could mean production code gets translated and sent to designers for the final tweaks before being fully integrated into the customer experience.
2. Democratizing website design
Companies like Wix build design principles into their website algorithms to help customers publish rich, user-friendly websites regardless of their design background. Simply tell the Wix AI-powered assistant what type of site you want to create — an e-commerce site, for example, blog, portfolio, or business site — and you’re off.
Once you’ve established the general type of site, you’ll be asked some basic questions — business name, location, and theme. Choose one of the design options that’s then presented and you’re more or less done — all in about two minutes.
There are countless other AI-powered possibilities as well. Solutions like Impress.ly let users make a style, color, or navigation change and apply it sitewide. This, again, is something we’ve traditionally had to manage — and depending on the complexity of a site, cascading those changes can be extremely time consuming.
And while it takes a little more work, it’s hard to ignore Sacha by Firedrop.AI and its chatbot that manages any AI customization you need. If you want design changes, tell the chatbot — and expect the bot to share a few creative suggestions with you as well.
3. Generating something from nothing
AI is also incredibly easy to tap into on This Person Does Not Exist. I admit, I’ve fallen down the rabbit hole on this site more than a few times. But once you get your fill of clicking “Another” to see a new face — a face of someone who doesn’t actually exist — there’s a lot of value in this face-generating site.
Instead of using stock images, you can simply click “Another” to get another face of another person who doesn’t actually exist, eliminating legal and usage issues while giving users endless faces to choose from for their designs.
Similarly, UI Faces is now available as an Adobe XD plugin that integrates This Person Does Not Exist. With it, UI designers can generate avatars in projects without leaving XD. Simply select one or more shapes you want to fill with avatars, select filters, and get your ideal persona — or just click around to generate avatars at random.**
**Taking things a step further, generative design algorithms are helping researchers figure out how to build a logo by choosing a color. To achieve this, designers are using generative adversarial networks (GANs), two-part neural networks that use generators to create samples and discriminators that distinguish between generated samples and genuine, real-life samples.
A research project called LoGAN used this AI technique to craft logos from a dozen distinct colors. This illustrates the potential to create multiple design options instantly — which normally would require hours of painstaking work, much of which won’t ultimately be used.
4. Ensuring compliance
There’s also the compliance side of things, which can be a big piece of the ongoing design and creative experience. From regulations like GDPR to ADA compliance, it’s essential your brand stay in-step with the latest policies, or risk significant fines — or worse.
This is another area AI can assist with the ongoing heavy lifting. Brands like CognitiveScale have established augmented intelligence-based machine learning systems that automate and improve compliance processes.
Likewise, Australian AI-powered compliance company Red Marker scans websites and analyzes marketing content for compliance risk issues. By converting legal checklists into digital rules, Red Marker can effectively manage compliance, even as rules and regulations change. This, the company says, boosts productivity and mitigates risk. To date, major global enterprises including Westpac, Citi, and CBA are integrating the platform into their digital experiences.
5. Embracing the potential of AI-powered UI/UX design tools
These examples are just scratching the surface. AI and machine learning are well-positioned to change the way we think about and act on UI and UX design.
At Adobe, for example, we’re working on an Adobe XD feature that would enable UI components to automatically scale based on content. As a design author implements or authorizes content into a design or design prototype, components will scale automatically. This will eliminate a tremendous amount of work that, until now, designers have had to tackle manually.
Overall, though, we’re seeing tremendous advances — and I, for one, can’t wait to see what’s next and how it will impact me, my team, and the industry as a whole. By embracing AI to simplify and streamline common tasks, designers have more time to pursue their creative visions while arming themselves with the tools they need to quickly deliver the best user experience. And that’s always a win — for both customers and brands.
Learn more about how designers and industry insiders are using AI and machine learning to create bigger, better, and more robust design experiences.