The next-generation product designer

The duties of a digital designer have begun a major shift. The change is so big that about half of Western designers are worried about their future according to the Figma 2026 study. Given people’s natural resistance to change, this stress is understandable and well placed. It is unavoidable that large-scale adoption of AI tools and agentic colleagues will influence what designers do. However, we believe that it won’t make designers useless as long as designers can adapt to the new configuration of work.

In this article, we elaborate on four arguments about the future of the product design profession in the AI age:

1. Some tasks will diminish
2. New tasks will challenge the professional comfort zone
3. Humans will continue to master emotional and intentional design
4. Good taste and quick judgment are increasingly critical

 

This article is a follow-up article to the Jagged Frontier of Design AI published on Qvik blog.

The type of design jobs that AI will steal

Let’s admit that right now it seems evident that AI is about to take a large bite of UI design, prototyping, and UX research work. AI seems to be well-suited to tasks that involve a lot of nibbling around these types of details, whether those are UI elements described in code or pieces of text. It is hard to generalise a description where AI excels, but well-defined environments, multiple constraints, and repetitive tasks are all subject to diminish in designers’ calendars.

This means that in some tasks, designer effort will be radically reduced. There’ll be less time in Figma pushing pixels, less time writing first-draft copy, less time building prototypes from scratch. If those tasks happen to be your favourite part of the job, you will have more adjusting to do. The good news is that it seems that AI will also give you easy access to flex your muscles on different types of tasks altogether which can hopefully inspire you, but does call for an extension of your professional comfort zone.

New outputs: redefining your professional comfort zone

The Microsoft Work Index report has done a job in 2025 and 2026 by introducing new concepts that help to envision how future work is getting organized. In 2025, they laid out the model of a frontier company that was imagining how human and AI agents may organize themselves. This argument was elaborated in 2026. The emphasis is still on the organization. Unless the organization embraces the possibility of change, it doesn’t matter how much the individuals expand their potential, they will be trapped. However, it is acknowledged that AI can lift the ceiling of individual potential. It is necessary that designers take that ride. But where does it take us?

To us, it seems inevitable that designers must move closer to code. This means designers would increasingly work with code instead of Figma layouts and prototypes. It doesn’t mean that designers would be responsible for shipping production code on a daily basis, but that they would comfortably interact with actual codebases and be better informed about how the designs can come to life in code. Not all designers and not all the time, but many more and more frequent.

One example of such an already happening shift can be found at Notion, an American productivity SaaS software. In an interview of Max Schoening at Lenny’s Podcast, Max describes Notion designers starting to work with code in a restricted code playground. As their skills and confidence grew, and the AI tools became more capable, they are “contributing to the production codebase.”

We think this is the way most organizations and designers will move in the future. Right now, few think production code is an expected outcome of a designer’s work, but given how easy code generation has become, that might as well be the case in the future especially for the front-end components. Then we would look back amused to the days when business stakeholders gave UI designers the assignment to update web app copy, which a front-end developer then implemented in production.

Of course, the new constellation of tasks means that besides coding, designers will clock more time prompting, reviewing, and refining generated flows as those are coming in from AI agents and assistants.

At the core of human design: mastering feelings and intents

While AI systems have proven themselves capable in many design tasks, they have their limits. These boundaries can be called the jagged frontier of design AI and it means AI can fail miserably in areas that require holistic integration of (vague) information from multiple sources. Such as typical product conversations inside medium and large companies regarding what the product should do.

AI can annotate your meetings, but it doesn’t make better sense out of the conversation than humans or propose compromises between conflicting points of view. For designers, often the lone advocates of the user perspective, neutral conversation analysis is not even sufficient as we have understood that we must fight for the right to a better user experience in the cross-fire of business and technical constraints. AI can mimic the user point-of-view but caring for users sounds hollow coming from an AI.

That hollowness is the opening. Once production becomes cheap, and on most digital product surfaces it now is, the screen stops being the valuable artifact. Anyone with a prompt can generate a screen. What can’t be generated is the reason a particular screen should exist: which user, in which moment, under which constraint, getting which thing they couldn’t get before. That reason is the intent. And the intent is what the future designer authors.

For decades, designers wrote use cases — descriptions of what a system should let users do. The vocabulary fit a world where producing the system was the slow, expensive part and the description was the cheap front-end work that fed it. That world has flipped. Description is now the slow, expensive, valuable part; production is the downstream consequence. The use case was previously a brief for the implementation. The intent is a brief the implementation can be executed from: by humans, by AI, by anyone the team chooses.

But a brief is only the vehicle. What makes intent valuable isn’t the document. It’s the stance it carries. It says: this user matters; that constraint gets respected; this outcome is worth defending; that other one is a nice-to-have we’ll cut. AI can mimic the document. It cannot take the stance — stances cost something (political capital, taste, conviction), and AI has none of those to spend.

The most valuable artifact a designer now produces is not the Figma file or a piece of code. It is the intent those artifacts get judged by. And judgment is something that designers are getting ever more involved in thanks to AI.

From the design tinkering to practicing good taste and judgment

For few years now, the advances in generative AI image generation has spurred people, including us, to create and occasionally publish images that felt awesome and fun at the time being, but didn’t really stand the test of time. How long that time is between awe and regret, feels like a good starting point in defining what taste is. It is easy to know it when you see it (“oh my god, that was awkward”), especially in retrospect. But there is no looking in the rearview mirror in the middle of the usually hectic everyday product development work.

The decisions, many of them, have to be made at an increasingly rapid pace.

Taste is an almost magical designer skill many American product people hold in high esteem. It is something that must be honed and cultivated to be ready for instant application. The judgment calls suddenly become a more frequent part of the designer job than they did when the time to design output was longer. Here’s the tricky part, it is very difficult to pinpoint exactly what that means. Read, for instance, the following excerpt of Figma CEO Dylan Field’s definition (in the same context he also defines craft and point of view):

“Taste is about navigating the possibilities of what’s out there and having preferences that are really clear you can articulate”

While that reveals something about decision making and its defenses, it doesn’t satisfy all curiosity. There’s a fresh, complementary definition from Max Schoening of Notion at Lenny’s Podcast in which he describes taste as “given an idea, you can predict for certain ingroup are they going to like it or not.” This emphasizes the predictability of decision making in situations where we don’t have a good precedent of a particular idea and its response. Schoening personally does not defend the human excellence of taste as defined in this way, but it still fits the context.

Summary: the easier and harder job of the next generation designer

Having the capability to always do more in your domain and adjacent one, is an opportunity and a risk. A couple years into the AI now, we are already aware how the new work can entice us into working ever harder, multitasking more, and reaching beyond our professional boundaries – something we would traditionally have called doing someone else’s work (Ranganathan & Ye in HBR, 2026). For designers, who will, in our vision, be reaching to new professional domains (or maybe we just redefine domains), this means more demand for self-control to prevent from losing focus on their own work. If you can contribute increasingly more to development, product marketing, or even support, how to maintain focus and excel in your own domain too?

In fact, is there anything easier about the future work? Speaking about cognitive demands, probably not. What we gain in automation of routine work, we lose in increased demand in human decision making, which certainly increases in quantity and qualitatively in complexity. AI can eat away slack and redundancies in design craft, leaving designers with the most demanding and human tasks. On the other hand, this could be an opportunity to establish new modes of working and at least find new kind of inspiration.

For example, many professionals feel their working hours are mostly spend in endless meetings, which take up time from doing “actual work.” What if we can turn the meetings more into “actual work” as the grunt work can be executed by agents almost as we speak? It is up to designers and their leaders to drive new organizational design that doesn’t abandon the employee experience amidst the fast-paced change in how we work.