The new bottleneck is knowing what to build

For much of digital product development, making the thing was expensive. Designers created interfaces. Engineers translated them into working software. Teams debated technical feasibility and waited for development cycles.

AI-powered prototyping is beginning to loosen that constraint.

The result is a peculiar inversion: when almost anyone can make something quickly, the ability to make something becomes less differentiating. Knowing what deserves to be made becomes more valuable.

Nick argues that the competitive advantage isn’t simply velocity. It is the discipline to question what all that velocity produces.

“An advantage is our ability to iterate,” he said. It also requires “the humility to go back and test yourself, to make sure that what the algorithm put out is, in fact, what the user wanted.”

That humility matters because AI doesn’t eliminate bad assumptions. It can industrialize them.

A product team that misunderstands its customers can now turn that misunderstanding into prototypes, features and entire experiences faster than before. Imagine an electric bicycle pointed down the wrong street: the motor is impressive right up until you realize how much farther it has carried you from your destination.

Nick used a similar e-bike metaphor during the conversation. AI gives teams what can feel like superpowers, but “that doesn’t necessarily mean we’re going in the right direction.”

This is where strategic UX research becomes consequential.

UX research has to escape the button

There has always been a temptation to reduce UX research to usability: Can someone find the button? Does the navigation make sense? Can the user complete the task?

Those questions matter. But they assume the experience deserves to exist in the first place.

Lija raised precisely that problem: How do researchers avoid becoming trapped in tactical questions about whether a particular interaction works and instead ask whether it enables something customers actually need?

Nick pointed toward service and experience design, where researchers step back from individual screens and examine a person’s broader workflow.

A confusing button might be a problem. But perhaps the user shouldn’t have been forced onto that screen at all.

“You don’t see those until you take a step back,” Nick said, describing frustrations that emerge across an entire day or task rather than within a single interaction.

That distinction could become more important as AI accelerates production. If building gets cheaper, questioning becomes more valuable.

The future of UX research may therefore depend less on defending old processes and more on moving upstream—toward customer problems, product strategy and evidence-backed decisions.

Give AI the drudgery, not the relationship

There is another mistake lurking in discussions about AI and user research: treating every part of the research process as equally suitable for automation.

Nick draws a line.

He recalled the painstaking work researchers have traditionally done after interviews: transcription, indexing, tagging and parsing enormous amounts of qualitative information. There can be value in that slow process because researchers absorb nuance while reviewing the material. But there is also plenty of labor that language models can perform efficiently.

Those are sensible places for UX research automation.

The human conversation is different.

Nick described learning that some academic institutions were using AI agents for initial admissions conversations. The idea troubled him. An admissions interview can contain ambitions, fears and subtleties that are difficult to reduce to a clean data-processing exercise.

The same principle applies to customer research.

Human insight comes partly from what people say, but also from context: hesitation, contradiction, emotion and the unexpected story behind an answer. Researchers create conditions in which people feel comfortable sharing those things.

Nick’s preference is straightforward: “Whenever possible, I would default to how can we become more humanist versus how can we become more efficient?”

That doesn’t make him anti-AI. It makes his argument about AI in UX research more interesting.

Use the technology aggressively where it removes unnecessary work. Be far more cautious when efficiency requires simulating the person whose experience you are supposedly trying to understand.

Democratizing research doesn’t mean removing researchers

AI also complicates an old debate over research democratization.

As research tools become easier to use, product managers, designers and marketers can conduct more customer research themselves. To some UX researchers, that can feel like an erosion of professional territory.

Nick sees another possibility.

Researchers can stop acting primarily as gatekeepers of insights and start building systems that allow an organization to engage with customer evidence more directly.

He described research repositories and portals where leaders can query findings instead of waiting for a researcher to deliver another presentation. The goal is not merely to present an insight and watch it disappear into a slide deck.

“You are also training them so that they are feeling empowered themselves to be able to query and understand the research,” Nick said.

That changes the researcher’s role from keeper of the answers to architect of organizational understanding.

And it addresses one of the profession’s persistent frustrations: excellent research that receives appreciative nods from leadership and then has little effect on the roadmap.

Customer insights become more influential when they are woven into how decisions get made, rather than delivered ceremonially at the end.

The business value of getting the direction right

UX researchers have long struggled to quantify their contribution. Revenue belongs to many forces. So do retention, conversion and churn.

Nick cautions against claiming too much.

Instead, researchers can connect their work to the outcomes closest to the experience they studied: abandonment, conversion rates, churn and other measurable behaviors. They can also demonstrate how deeply research informs product and design decisions.

The business value of UX research isn’t only that it helps companies make good decisions. Sometimes it prevents expensive bad ones.

That proposition becomes more significant when AI allows organizations to execute those bad decisions faster.

Research, in that sense, is navigation rather than a brake. Its job isn’t to slow down the machine. It is to make speed useful.

The most human job in the room

Researchers understandably wonder what all this means for their careers.

Nick doesn’t promise immunity from technological change. Designers, engineers, product managers and researchers are all watching AI absorb tasks that once belonged exclusively to them.

His answer is not to retreat from AI but to learn it—and then double down on what the technology makes more valuable.

Researchers understand people. At their best, they connect customers with designers, engineers, product managers and executives. They facilitate difficult conversations, challenge assumptions and translate messy human behavior into something organizations can act on.

“You, the humanist, are the glue between these teams,” Nick said. “You are the one that’s supposed to understand human beings better than anyone.”

That may be the most compelling argument for the future of UX research.

AI can make prototypes faster. It can summarize transcripts, organize customer insights and help researchers move through enormous amounts of information. Researchers should use those capabilities rather than pretend they don’t exist.

But efficiency is not empathy, and information is not understanding.

The opportunity is to use AI to create more room for the work that requires judgment, curiosity and human connection—the work of figuring out not merely how to build something, but why it deserves to be built.

As Nick put it: “What we’re holding on to is that conversation, is that connection because nothing substitutes that.”

Related resources and links

• Evolve your UX research strategy to win in an AI world. This on-demand webinar examines how UX teams can use AI to accelerate research and synthesis while preserving human judgment, and how researchers can scale insight without sacrificing confidence or quality.
• UX Research Leadership Playbook: Scaling Insight in the Age of AI. This playbook examines how UX research leaders can evolve into strategic business partners by scaling customer insights with AI while maintaining governance, quality, and human-centered decision-making.
• AI in UX research with John Whalen. This Insights Unlocked episode explores many of the same tensions as Nick’s interview, including AI moderation, simulated users, AI research tools, and the importance of using AI to augment rather than replace human insight.
• AI might finally deliver on the democratisation of UX research—but only with researchers at the centre. This blog post argues that AI can enable designers and product managers to handle more tactical research while freeing UX researchers to tackle strategic questions—provided researchers establish the right quality and governance guardrails.

• Nick on LinkedIn

• Lija on LinkedIn 

• Nick's personal website

• Gauge.io

• Nathan on LinkedIn