This week's UXD <> AI signals worth sharing

3️⃣ UXD <> AI news stories

• AI integration in UX design is improving UX, says a report.

• Pentagram used AI to design 1,500 icons for a federal govt. website.

• Figma’s Config conference demoed massive AI-powered features.

1️⃣ Helpful AI tool for UXD

• Flowy: a UX <> AI tool created and developed by students at Cornell University who identified a gap in other UX/AI tools.

1️⃣ UXD prompt of the week

• 🎯 Prompt of the week: Rewrite as 3 specific hypotheses I can test as a designer.

3️⃣ UXD <> AI news stories

AI integration in UX design workflows is improving UX, says a report.

AI Integration in UX Design: recent discussions highlight how AI is transforming UX design by enabling features like screenshot-based actions, gesture-triggered AI responses, and dynamic UI generation. These advancements are redefining user interactions and setting new standards for intuitive design.

A fresh research report titled “AI and the Future of Design” dropped this week, synthesizing industry surveys, case studies, and expert commentary. The report is behind a paywall but you can see a preview here. One big takeaway: while AI boosts efficiency in prototyping, research, and iteration, human judgment is still irreplaceable when it comes to ethics, context, and nuance. The report pushes designers to get fluent in AI not to compete with it, but to shape how it's used.

It’ll be interesting to see how this plays out if AI features do indeed improve UX however I’m skeptical when corporations are at the helm. They are notorious from being shut off from reality and make decisions based upon that willful isolation - we see it in every industry from auto to entertainment to energy. Think about it, folks at the executive level have been there for a long time and many of them will turn a corner to become so out of touch with reality. So how can you influence UX when you no longer understand who you’re trying to influence? We’ve seen this on a continual loop for longer than I’ve been alive. I’m not convinced the tech industry is going to be any different. Regardless the report is fascinating and it’s worth a peek.

Pentagram used AI to design 1,500 icons for a federal govt. website stirring up questions about scale, clarity, and the future of craft.

In a headline-grabbing move, design agency Pentagram leaned on Midjourney to generate 1,500 icons for the U.S. government’s Performance.gov. While the team defended it as a way to explore visual systems and work efficiently, the design community split: does this level of automation dilute meaning? Are we prioritizing quantity over clarity? It’s a big swing and a big case study in how AI can help scale, but not always sensibly.

As someone who uses AI as part of my everyday workflow, I don’t see any issues with using AI to speed up work especially for an enormous work load like creating 1,500 icons. As a designer and an artist (I’m gonna give my painting website a plug) though, I think these icons are muddy and distasteful. It is not a strong execution and they’re just kind of gross. This is a fault of the prompt engineers and human creative direction, not Midjourney or AI. It’s too bad that it worked out this way because it could have been an interesting public dialogue about AI, efficiency and human creative direction in a more positive way.

Figma’s Config conference demoed massive AI-powered features.

At this year’s Figma Config, all eyes were on the platform’s expanding AI capabilities. Expect features that let you generate UI from text prompts, auto-fill design systems, and co-edit with an AI assistant that “understands your intent.”. AI as your pair designer. But it also sparked a broader convo: Are we designing less and directing more? And if so, what new skills will designers need to stay relevant? Is Figma’s future vision of designers that we are prompt engineers? I would guess they would say no. Maybe next year they’ll focus on the human creative direction in designing with AI.

I have so many thoughts about this week’s Config and maybe I’ll do a seperate newsletter to unpack everything. Here’s my three biggest takeaways:

1. I’m excited to check out all the new tools and play around with them. Especially Figma Draw and Figma Sites. This might be a turning point in AI and UXD where there’s solid tools to help us expedite our design workflows and create more meaningful work.

2. I’m thrilled that they did some serious housekeeping on existing tasks and minimized steps to execution. My biggest beef with Figma has always been that it’s too cumbersome to do even the smallest of tasks and that results in poor design driven by weak thinking because the tool is too difficult to start anything new. A couple of the tasks they sped up are auto-layout (!!! the speed - I love it!) and the interactive behaviors, like the super easy and fast, hover effect to add to buttons, etc!

3. Piggy backing off of my point number 1: I’m excited to use the new tools and when I watched the Figma Sites keynote, the first thing I thought is I might migrate my painting website there as it’s time for a re-do.

   Then I saw the keynote with Meta and Figma CEO Dylan Fields on Building the next computing platform where they kicked off talking about defining consciousness and basically said that we might just have to accept that humans will not be the sole owners of consciousness in the future.

   First off, we’re not as all animals have consciousness too.

   Second off, this public dialogue on AI and the machines overcoming humans is filled with so many assumptions about someone’s perceived interpretation of “reality”. There’s a lot of talk about “once AI learns morality humans will be overtaken”. That’s operating on an assumption that morality is the driving force of humans. I’m not convinced that morality is a driving force in anyone. I think it might be about preservation of self, species and the ability to grow both of those things. If my assumption has some degree of accuracy to it, could AI ever even learn that concept? I don’t know and we’ll find out as time goes on. For right now though, I’m not going to worry about a future if neither myself nor anyone else has enough data points to accurately assess the risk factor.

   Anyway back to this keynote, it kind of reminds of Curtis Yarvin, The Network State and the technoautocratic state. If you don’t know about it, look it up as that’s too much to jump into here.

   Anyway, Meta also just started defense contracts in October 2024 and their only product is our data. Yes I still use Meta products out of necessity however if I can cut out any amount of my data that is going to be used in a defense contract then I will. Figma is easy to cut out.

   I do not know the extent of this partnership and if Figma data is be used in conjunction with Meta’s defense contracts. Regardless, this partnership with Meta is a huge red flag and my alarm bells are going off.

1️⃣ Helpful AI tool for UXD

Flowy: a UX <> AI tool created and developed by students at Cornell University who identified a gap in other UX/AI tools: “they overlook the crucial aspect of interactions and user experiences across multiple screens.”

Flowy: A new AI-powered tool designed to assist UX designers by annotating multi-screen user flows with relevant design patterns. It leverages large multimodal AI models to enhance the ideation process, helping designers identify and understand abstract design patterns across multiple screens.

This tool is pretty cool and seems to have solved a real gap in UX <> AI tools.

1️⃣ Prompt of the week

🎯 Prompt of the week: Rewrite <a vague request> as 3 specific hypotheses I can test as a designer.

This is one of my favorite prompts Ask ChatGPT to rewrite a vague request you’ve recently received, like ‘make it better’ or ‘clean up the UX’ as 3 specific hypotheses you could test as a designer. The results will put you on a good path.

This prompt begins the arduous process of assumption gathering and clears up confusion at the onset of a request that can derail the entire team.

That’s it for now. Linda ✨