UX ecosystems in the age of AI

From The Inside Seat Edition 1: A conversation with Clara Kliman-Silver on how design systems, tools, and people have transformed in the age of AI

We caught up with Clara Kliman-Silver, Staff UX Researcher at Google, who focuses on how people build products and measure productivity. She specializes in participatory design and generative methods to investigate workflows and understand designer-developer experiences. You may have seen her speak at UX New Zealand on artificial intelligence & design: imagining the future of UX back in 2023.

We caught up to discuss how the UX ecosystem of a design system, tools, and people has transformed in the age of AI. Clara shared her personal take on what it means for designers, researchers, and teams navigating this shift.

How has the UX ecosystem evolved since you last spoke about it 3 years ago?

The biggest thing is that the tools have changed. We’re having this age of AI-driven creation tools, analysis tools, development tools. But I think at the heart of it, you still want design consistency, which is what a design system can bring. You still want usability, like, good usability. You still want all the rules of visual perception and interactions still apply. So I think it’s really more how has the mechanism through which we communicate or shape the experience has changed.

The fundamentals like design consistency, good usability, and the rules of visual perception still apply. What’s shifted is the democratization of some fields.

With the advent of generative AI tools suddenly, specialties have really opened up. So, someone who isn’t a web developer or an app developer can vibe code their way to an app. And we see the same thing for design, and for research, especially. There are a lot of tools that have helped democratize some of these fields.

Is UX dead?

We’ve all seen the headlines: UX is dead, followed immediately by arguments that researchers are more important than ever. The tension in the discussions about the future of UX is real.

I think the way that we do this work is going to change. I think the roles themselves may constitute different things, but I see it more of an evolution and a parallelization. Maybe you can run more studies or conduct more experiments in parallel but that means that you as a UXR need to level up and be able to think at a higher level. If I had all this bandwidth to be able to run all these things in parallel, what could I do? But then also, you might spend more time figuring out what do the results mean, how do I pitch them to the right people, how do I extrapolate, what other kinds of testing might I need to do to figure out where to go from here.

Roles are evolving. Where and how you spend your time is evolving.

Yes, there is this question of fragility in the background, but I do think that it’s also a question of opportunity, and sort of the roles will change, and new capabilities will be unlocked as a result.

AI is changing the game.

I think the culminating thing is that it’s allowing people to expand their roles or deepen their skills in ways they couldn’t previously.

The real story is opportunity.

You’re also seeing this in the blending of roles, like, more developers taking on design work, more non-software engineers or non-coders taking on development work. I think that it’s facilitating a new kind of collaboration, and with that comes new kind of innovation, because you have new people and new capabilities, they’re able to come up with novel concepts.

What’s the biggest thing that changes with this shift?

Perhaps two things have changed. One of them is you have all these fresh perspectives and new people coming into these spaces, which I think is generally a really good thing. Secondly, at the same time, that means that it’s ever more important to have a human in the loop.

It’s about staying grounded outside of the AI output with real human partners so you can identify unexpected outcomes and use their judgement and expertise to intervene.

If something is about to go out and affect millions of people, or literally lead to life or death decisions, that you really need to make sure that the risk is non-existent, or at least as low as possible. I think the first question to ask is what is it that you’re trying to do? What is the most important thing to do here? Then to think about, okay, who and or what can do it?

The ideal process

For me, the ideal process is we do the study, and then within a few days we turn around a list of insights we think are most important and what we think the team should do about it.

The important note to highlight is early dialogue between functions in discovery. What stands out to a researcher might not be a big deal from a technical perspective. Or someone with different expertise might catch a larger problem because of their expertise in that domain.

Co-creation can be really helpful in getting the team to move quickly after you do a study, but also really to have them be part of the discovery process and help you identify where there might be things that you’re missing that are still really important for a good user experience.

How can cross-functional teams navigate this shift?

Amidst hyper-optimization and scramble, break down what you’re really trying to do.

First of all, what is it that you’re trying to do in the first place? And what are the things that must happen? Again, regardless of what tool or technology is used to complete them, or who’s responsible, but, what is the minimum viable process or success look like here?

Then, get clear on what productivity actually means for your team.

Then also ask, what does it mean to be more productive? I think that sometimes people say, well, I want to be more productive, which just means faster output, and faster output is a great thing, if that’s important, but sometimes faster output means lower quality. So then the question is, again, what are the risks that you’re willing to take to work more quickly, or to be more thorough?

New tools require adjustment time. Procedural changes, workflow shifts, how information gets stored and accessed, the quirks of each tool, all of it takes time to learn and has ripple effects across the team.

Obviously, seize the moment, see what’s out there, see what you can use that will help you, see what new capabilities you and your teammates can take on. But be aware of what might not be realized as productivity gains right away. And in fact, you might have, in some cases, slower velocity for a while you’re going through this change, and then you’ll see the real benefits later. And also, don’t be afraid to try something else. There’s a lot of really good and useful things out there, but there’s also a lot of hype.

Remember, context is everything. When you see success stories online about tools changing everything, remember they’re speaking from their specific circumstances. A startup building from scratch operates differently than a team scaling a 20-year-old codebase.

So when you’re going online and you’re seeing people say, “this tool or technology completely changed how I do X, Y, Z,” keep that in mind. What are your specific circumstances, and what are the accelerators likely to be in those cases?

Keep measuring and improving your practice, together.

There are a lot of well-established tools and approaches to understand team effectiveness.

Don’t go reinvent the wheel. There’s a lot of really good stuff that’s been out there and tested over long periods of time, and just because something is new and available doesn’t necessarily mean it’s better than what came before, or it has to operate fully on its own.

Retrospectives continue to shine to track what was successful, what keeps coming up, and where are the opportunities for the next go.

Continue to do your retros. Lots of agile software development among other models, have some really nice built-in retrospective moments where teams can come together and say, what went well last time? What do we want to change? What is our start, stop, continue model?

The foundations that have always mattered are still important.

We are in a really exciting time where there are a lot of things changing and a lot of new things available to us, but there are also these frameworks and underpinnings that are constant, or are not changing as quickly, and I think it’s important to remember that, and to go back to those whenever we can.

🌐 Head to our LinkedIn to watch the conversation, and follow for more candid conversations on UX in the age of AI.