Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Did AI kill the research star?

Research should help shape business decisions, and humans play an important role in that.

[Photo: Getty Images]

Profile photo of Jason Giles

VP of Customer Intelligence · Auros

AI is changing some of the work researchers do today. And for many, that’s an uncomfortable reality. It can summarize interviews, synthesize data, generate personas, and produce polished reports faster than researchers can manually. I don’t think the future of customer and user research is about competing with an algorithm to see who can synthesize a transcript more quickly.

AI has exposed a flawed assumption that the value of research is in the output it produces. The real value is helping the business make a defensible decision.

AI IS LOWERING BARRIERS

We’ve been through a version of this conversation before. As customer and user research tools became more accessible over the past decade, more people could conduct research themselves, democratizing that ability. Researchers worried that a lack of formal training could compromise research quality and rigor. With the right structures in place, I’d argue that those concerns didn’t fully materialize.

Democratization lowered the barrier to doing research. AI has lowered the barrier to producing conclusions. Anyone can ask AI to analyze information and return something polished and convincing. But convincing isn’t the same as correct.

AI can be a phenomenal synthesis tool. The problem comes when we treat a convincing output as a substitute for evidence, particularly when a decision has real consequences for customers or the business.

That’s where the researcher’s role needs to change.

THE COST OF CONSEQUENCES

AI has made producing and revising work much cheaper. That can make it tempting to assume that getting something wrong has become cheaper, too. It hasn’t.

Changing a design file is cheap. Shipping the wrong experience and losing revenue because of it isn’t.

As AI removes the natural speed limit of human production, organizations can generate more concepts, conclusions, and decisions. Accountability remains. Human judgment becomes even more important as ambiguity and the consequences of getting a decision wrong increase.

Researchers can help leaders calibrate where a decision falls along that spectrum and where the evidence needs to be airtight.

PUT HUMAN JUDGMENT WHERE THE RISK IS

I think about this as a risk gradient.

For low-risk, reversible decisions, AI acceleration is a clear advantage. Let the machine draft routine copy, summarize data, and help with early exploration. As decisions become more consequential and difficult to reverse, careful human judgment and empirical interpretation become more important.

That’s why “human in the loop” isn’t always a sufficient safety net. A human reviewing a polished AI output doesn’t necessarily have what they need to determine whether the conclusion is right.

For high-stakes decisions, we need empirical evidence in the loop, too.

FROM METHODS EXPERT TO STRATEGIC NAVIGATOR

This is the bigger shift I think researchers need to make. AI’s value is in serving as a neutral summarizer. Human researchers should be strategic navigators. Their value is in the ability to understand messy organizational context and calibrate risk. They understand methodological rigor. Human researchers can articulate options, prioritize them, and make recommendations even when the evidence is incomplete. This allows leaders to make defensible decisions.

This came up recently when I asked a trusted researcher to help us develop our company’s point of view on AI moderation. We wanted to understand when it makes sense for AI to lead a research conversation, and where human judgment is still needed. In this case, AI would take the place of the human moderator, speaking with a human participant, responding to their answers, and adjusting its questions as the conversation unfolds. The researcher brought me data and findings from active research, but she also understood our brand, our company’s history, and the principles we operate by. That broader perspective helped us determine how to frame the issue and ultimately, how to apply the technology. Her contribution went well beyond what the facts and customer feedback alone could tell us.

WHAT I NEED FROM RESEARCHERS IS CHANGING

Think of AI as a car traveling 100 miles an hour down a multi-lane highway. I don’t need my research team to ride the brakes or keep the car in the garage. I need them to be the navigator, helping us understand where we’re going, when to change course, and when the road ahead requires more caution. That responsibility doesn’t sit with the research team alone. As AI accelerates how quickly organizations can produce and act, leaders also need to understand when research should guide a decision rather than be bypassed in the name of speed.

So, did AI kill the research star? No. But it has made research theater harder to defend. A polished report has never been the point of research. The point is helping a business make a sound decision.

AI can move us from raw information to a plausible answer in seconds. That answer still needs to be tested against customer evidence, business context, and the consequences of getting it wrong. Researchers bring that judgment to the process. They help leaders understand what the evidence supports, where uncertainty remains, and how to move forward.

AI can help us move faster, while researchers help us know where we’re going.

Jason Giles is VP of customer intelligence at UserTesting.