The Bull And Bear Case For Digital Design In The Age Of AI — Smashing Magazine

Design’s Real AI Shift: Permission, Not Pixels

For years, designers have made the same quiet complaint: they could do better work if the organization got out of the way. The specific phrasing changes — not enough engineering time, product already fixed the solution, the roadmap was too full, research got cut — but the underlying frustration is consistent. Designers sit in an awkward middle space. Product frames the problem, engineering decides what is affordable, and design is expected to make the result clear, coherent and usable without disrupting anyone’s plans. They are told to think strategically while lacking the power to act strategically.

Designers can spot the broken onboarding flow, the confusing upgrade path, the empty state that makes users feel stupid. Seeing the problem is one thing. Getting it fixed is another.

That gap turns a lot of design work into an exercise in persuasion. You build the case with research clips, support tickets and polished prototypes. You explain why the “small edge case” is actually the first-run experience for half your new users. Everyone nods and moves on to the roadmap they already committed to.

This context explains why AI matters more to design than the usual “will it replace designers?” debate. The interesting change is not that designers can now produce more screens — nobody needs more screens. The shift is that designers may need less permission to make changes real.

The Bull Case: Making Ideas Hard To Ignore

With modern AI tools, a capable designer can move from “we should fix this” to “I fixed this and pushed it live.” They can prototype an alternative onboarding flow, write and test clearer copy, build a rough working version of an interaction, and clear small pieces of design debt without waiting for a roadmap slot. They can make the better version visible enough that it becomes harder to ignore.

That shift changes the politics of the job. Design has historically leaned on argument because designers lacked direct means of production. AI weakens that dependency. Complex products still carry architecture, data models, permissions, security, compliance and legacy systems, but the boundary is moving. A motivated designer with the right tools can now cross more of the gap between having an idea and making that idea real.

In this future, designers become less permission-dependent — less reliant on product to bless the problem, less reliant on engineering to make every small improvement happen, and more able to make, test, repair and ship. The strongest practitioners start to resemble hybrid product leaders. They still care about interaction, hierarchy, language, flow and craft, but they also understand the commercial shape of the problem. They can prototype in code, or close enough. They can use AI to explore options quickly and apply judgment to discard most of them.

There may be fewer of these people, but they will be harder to ignore. Much of the current design-org model grew up around scarcity: scarce engineering time, slow production, expensive prototypes and heavy coordination. If AI reduces that scarcity, it will probably shrink the roles built around it. The believable version of this outcome is not that every designer keeps their job with a productivity boost. It is that total designer headcount drops while the remaining designers hold more direct influence over the product. For the strongest in the field, that may be exactly what they have wanted for years.

The Bear Case: Autonomy Has Teeth

Autonomy also removes cover. The constraints that held good designers back also protected weaker ones from being tested directly. For years, it was easy to claim a better idea never got engineering time. Sometimes true. Sometimes the better idea was only ever a critique — never made concrete, never tested, never forced to deal with awkward trade-offs. It sounded strong because it lived safely in opposition to the shipped product.

A lot of designers are good at noticing what is wrong. Fewer are good at deciding what should happen instead. Fewer still can make that alternative real enough for other people to judge.

AI exposes that gap. If you can prototype the recommendation, the recommendation has to hold up. If you can build the alternative flow, the flow has to survive contact with details. If you can test the product copy, you have to care about what users actually do with it. Some designers have learned the language of strategy without the discomfort of owning outcomes. They can talk fluently about user needs and systems thinking but struggle when forced to make a call. They want influence without the exposure that comes with it.

The profession has spent years arguing that design deserves more power. More power means fewer excuses. Work gets judged less on the elegance of the argument and more on the quality of the thing made, tested or changed. That is a better standard, but it will not be kind to everyone.

The Second Bear Case: Plausible Design Ships

There is a more worrying scenario for large design teams. In most companies, product and engineering already hold more institutional power than design. They own the roadmap, the technical architecture, the metrics and the language leadership understands. AI may not rebalance that power — it may hand product and engineering enough design capability to make design easier to bypass entirely.

A PM who can generate a decent flow, decent copy and a decent prototype may not feel the need to involve design early. An engineer who can produce a reasonable interface may decide the design system handles enough of the decision-making. A founder who can reach a polished demo in an afternoon may confuse polish with product thinking. The risk is not that these people suddenly become great designers. The risk is that many companies cannot distinguish great design from plausible design.

Plausible design looks coherent in a product review. It uses the right components. The spacing is fine. The copy is not embarrassing. Nobody in the meeting feels strongly enough to object. So it ships.

Bad product decisions already survive because they look plausible. AI will produce more of them. Design loses ground not because taste, judgment and research stop mattering, but because the visible outputs of design become easier for other functions to imitate. If a company already thinks design is mostly screens, prototypes and polish, AI gives it a cheaper way to get those things.

In this world, design does not gain agency. It gets narrowed. Remaining designers manage the design system, police component usage, review flows that were already decided, tidy interfaces and get pulled into high-stakes launches. Useful work, but a smaller surface area — less shaping the product, more maintaining the furniture.

This is why the “AI will automate the boring 20%” argument feels too comforting. In some companies, perhaps that is what happens. But in large tech organisations where design teams grew around coordination, production and process, the cut could be much deeper — not 20%, but 50% or more, especially where leadership never understood why the design team had grown so large in the first place.

Two Futures, One Profession

The uncomfortable truth is that the most optimistic and pessimistic forecasts for designers in an AI-driven industry can both play out. AI can amplify the capabilities of the best designers while simultaneously making many average ones redundant. It can grant design more influence over products while shrinking the number of designers needed to achieve it. A few designers may use AI to climb toward product leadership, while others find themselves pushed into governance, cleanup, and maintenance roles. Some will discover that AI removes the need for permission, only to realize they were more comfortable waiting than acting.

The designers who do well will not be the ones who merely use AI to produce more options. Options are cheap now. They will be the ones who know which option is worth pursuing, why it matters, how to test it, what to cut, where the product is lying to itself, and when “good enough” is quietly damaging the business.

“

This shifts the burden from production to discernment. Designers will need product judgment, technical curiosity, commercial awareness, and the confidence to commit before all variables are resolved. The role will demand comfort in moving between a customer conversation, a prototype, a pricing concern, a brand question, and a messy implementation detail, without insisting that responsibility for each belongs elsewhere.

The resolution is likely to be an uncomfortable mix of both extremes. Some designers will secure more agency through AI; some companies will use it to require fewer designers. Some teams will shorten the distance between judgment and execution, producing better work. Others will ship more plausible mediocrity because no one in the room can distinguish it from quality. Designers have long argued that they could deliver more value with fewer organizational constraints. AI is about to test that claim, and the results will be revealing.

Further Reading

• “Good from Afar, But Far from Good: AI Prototyping in Real Design Contexts,” Huei-Hsin Wang and Megan Brown (NN/Group)
  A review of AI-powered prototyping tools covering real design scenarios found that while these tools can follow instructions toward a general goal, they often lack the sophistication to weigh design tradeoffs and produce thoughtful, high-quality outputs without extensive human guidance.
• “AI Design Tools Are Marginally Better: Status Update,” Megan Brown, Caleb Sponheim and Taylor Dykes (NN/Group)
  AI design tools have improved but still fall short of their promise. This article evaluates features including Figma’s Rename Layers, Rewrite This, Find More Like, Khroma Color, Midjourney, and wireframe and prototype generation tools.
• “Using AI for UX Work: Study Guide,” Tanner Kohler (NN/Group)
  A curated collection of links covering practical ways to introduce artificial intelligence into UX design work.
• “I used AI for every task for two weeks,” Joanna Otmianowska (DEV Community)
  A front-end developer recounts a two-week experiment using Claude Code for every work task, sharing the detailed results of the experiment.
• “How AI will Affect the Design Industry,” Andy Budd
  AI is unlikely to “kill design” in the near term, but the industry is in a phase of significant change that will create sizable opportunities for early adopters.
• “Design has been too settled for too long,” Andy Budd
  Despite its focus on change, the design discipline operates on a surprisingly fixed model. AI is beginning to break that model, with a detailed look at emerging trends in how design teams adopt AI daily.
• “What Designers Should Take From Benedict Evans’ Latest AI Deck,” Andy Budd
  Benedict Evans’ strategy decks offer a useful weather map for the technology industry: where money flows, what assumptions people make, which comparisons are lazy, and where the industry may be fooling itself.
• design + AI conference