Big Tech Has Tried to Steal My Job for Years. So What?
I test drove an AI interface design tool to see if it could replace the workflow I've taught my students for years. What I found should resonate with instructors in every discipline.
Like many instructors, I’m faced with big questions about the impact of AI on how I teach and assess learning. These questions include:
Should I still require students to create something from scratch? If so, why? And how do I convince students there is value in this “old-school” process?
Should I let students take every shortcut available? If so, how can I raise my expectations and adjust my rubrics to focus on the skills AI can’t replace (yet)?
How do I reassure my students that there’s still a need for the skills I’ve spent decades mastering? And how do I prepare them for a rapidly changing job market when I feel overwhelmed by these changes myself?
As part of my effort to answer these questions, I'm currently testing AI tools that my UI/UX design students might use to generate app interface prototypes. In the courses I teach, students typically use a tool called Figma to design mockups of app and website interfaces. Figma is one of the most popular interface-design tools on the market, and it supports variety of new AI plugins and features. I haven’t had a chance to test many of those yet, but I recently came across a tool called Uizard, which bills itself as a simpler alternative to Figma with a “lower learning curve.” It offers a free version with a limited number of projects, so I decided to give it a 60-second test drive.
I started by entering the prompt, "Create a gym app to book pickleball classes in the style of Airbnb" and clicked the visual-style preferences for "dark," "young," and "modern." In seconds, Uizard generated an interactive prototype with a shareable link.
You can view the interactive prototype here and browse screenshots below.
What Worked Well
✔ Shapes and text layers are fully editable. (This was a stark contrast from another tool I tried, which only generated non-editable PNG files.)
✔ Icons are rendered as vectors, which means they can be scaled to any size without a loss of resolution.
✔ Specific interface elements are linked in a (relatively) logical way, making the prototype functional right out of the gate.
✔ Spacing, alignment, visual hierarchy, and "scannability" are solid overall. This is a major challenge for my 100-level students as they work on their attention to detail. (It’s easy for inconsistencies to pop up, especially as a project grows.)
What Needs Improvement
✖ Parts of the navigation flow don't make sense. For example: clicking "book class" takes me to a screen with my stats and a leaderboard.
✖ Longer button text spills out of the shapes surrounding it.
✖ A decorative image on the homepage had garbled text. This is a common issue with the last generation of AI image generators. (They struggle to render readable text.) However, this has improved significantly in recent months, so there are plenty of workarounds if a student takes the time to use a separate tool to generate these images.
✖ The design feels flat and generic overall and could benefit from some experimentation with texture, depth, and color.
Impact on Assignment Design
On the one hand, I empathize with instructors who are weary of incorporating AI into their assignments. There’s an understandable sense of existential dread—do I stand firm and fight to pass down my values to the next generation? Or do I give in on these shortcuts I feel pressured to accept? Do I try to convince them that Scrabble is still the coolest game around, or hand over my iPad and enjoy a nice, quiet dinner with someone who still appreciates my love of a tactile triple word score?
I get why these feel like our only options at times, but there are plenty of shades of gray to be explored. The fact that Scrabble is still sold in stores (and that stores still exist) is proof that the old and the new can coexist. That coexistence isn’t always peaceful—but it’s possible, and that gives me hope.
Personally, I’m not too concerned about the impact of AI on art and design, and I plan to encourage my students to explore the good, the bad, and the ugly that AI-driven design tools have to offer. If I find that AI lets them do something faster than they could before, I have plenty of ideas for new learning objectives and assignments that would’ve been too ambitious in the past.
Dreaming a Bigger Dream for Students
In a 2016 Vulture interview, RuPaul was asked what he thinks about shows like Lip Sync Battle, where straight people perform competitive lip syncs to popular songs much like drag queens have done for years on his show, Drag Race. I’ve been thinking lately about his reply:
"Oh, I don't think of it. It's a poor ripoff of our show. Regular, straight pop culture has liberally lifted things from gay culture as long as I can remember. And that's fine, because guess what? We have so much more where that comes from. Take it!"
If I have but one bit of wisdom to share with every anxious instructor out there, it would be this: AI can impersonate you all it wants, but it can’t replace you. Britney and Adele headline at the biggest venues in Vegas while impersonators fight for scraps off the strip. Let’s dream a bigger dream for our students than impersonation.
