What Bitcoin Can Teach Us About AI-Generated Qualitative Research
The Problem with a Standard of Perfection
If you've spent any time in the world of customer research, you've probably heard some version of this refrain:
“We can’t trust AI to do qualitative research—it hallucinates. It makes things up.”
To which I say: Have you ever done real qualitative research?
I don’t ask that sarcastically. I mean it genuinely. Because if you have, then you already know this truth:
Qualitative research has always been messy.
And it has always relied on a fragile ecosystem of planning, execution, and interpretation. Let’s take a look at the real world for a moment:
Sometimes your sample isn’t representative.
Sometimes customers don’t say much.
Sometimes they say the “right things” but don’t mean them.
Sometimes the moderator leads the witness.
Sometimes insights are cherry-picked or retrofitted.
And sometimes, we simply get it wrong.
Yet no one is out there canceling focus groups.
The Flawed Standard
The problem isn’t that AI makes mistakes. The problem is that critics compare it to a fantasy version of traditional research—one where every respondent is articulate, every insight is perfectly valid, and the moderator is a robot trained in ethnographic purity.
That’s not research. That’s mythology.
And it’s a move we’ve seen before.
The Bitcoin Analogy
If this dynamic feels familiar, it’s because we’ve seen it in another domain that’s often misunderstood: Bitcoin.
Critics of Bitcoin will say:
“It consumes too much energy.”
“It’s based on nothing but belief.”
“It’s volatile.”
“It has no intrinsic value.”
Some of that is fair, but the real problem is the baseline they’re comparing it to.
They’re not comparing Bitcoin to the messy reality of the existing system.
They’re comparing it to a flawless ideal of a stable, trustworthy currency.
Most often, that currency is the US Dollar.
Let’s interrogate that for a moment.
How Stable Is the Dollar, Really?
In my lifetime, the US Dollar has lost about 90% of its purchasing power, according to the CPI.
But if you dig deeper, you’ll find that the CPI has been “adjusted” over time to soften the appearance of inflation. If you’re the skeptical type (and I am), you might raise an eyebrow at that. Wink wink. ;)
Compare the Dollar to gold? It’s down 99%.
This has been done purposely, by Keynesian economists, the system taught in every business school in the US. Just in the last five years, the U.S. has increased it 4x. (And you can expect this increase to accelerate… but I’m getting a bit off topic.)
So when people say, “Bitcoin is too unstable,” I ask: Compared to what?
Because here’s the thing—yes, Bitcoin consumes a lot of energy.
But that energy secures the network. It aligns incentives. It enables decentralization.
And yes, it requires faith. But so does everything else of value—real estate, wine, baseball cards… even the Dollar. Perhaps, especially the Dollar.
Can we trust AI Customer Research?
AGAIN I SAY: COMPARED TO WHAT?
AI-generated qualitative research has flaws. No doubt.
But those flaws are visible, often auditable, and improving fast.
And the value it provides—speed, scale, cost-efficiency, pattern recognition—is very real.
The skeptics hold AI research to an impossible standard. They say:
“It hallucinates.”
“It might be wrong.”
“We don’t know where it came from.”
But again: Compared to what?
Because I’ve seen traditional research processes with biases, small sample sizes, poor note-taking, and zero consistency between moderators. I’ve seen expensive studies that couldn’t be replicated or audited.
At least with AI, we have a transcript. We can examine the prompt. We can tune the model.
And we can generate 50 interviews in 5 minutes to pressure-test a hypothesis.
Is it perfect? No.
But neither is the alternative.
A Note to Practitioners and Leaders
I’m not saying to blindly trust AI. In fact, don’t.
I’m saying: trust it like you trust everything else in the research process—with critical thinking, context, and healthy skepticism.
But don’t reject it just because it doesn’t reach a mythical standard that your current methods also fail to meet.
Do I think AI research is “better” than human research?
It depends on the criteria.
And for teams with limited budgets, time, or research muscle—AI may not be a compromise at all. It might be the only way they learn anything.
AND HERE IS A POINT WHERE I’VE ALWAYS BEEN A CONTRARIAN:
I have never trusted any single study. I’ve always recommended triangulating from with various data sources. Triangulating to the truth. The patterns are there.
One More Thing
As I write this, Bitcoin has hit its second all-time high this week.
Which means: everyone who has ever bought Bitcoin and held it is in profit.
Meanwhile, everyone who’s held their savings in a US bank account has lost purchasing power—even if the number on the screen looks stable.
Think about that next time someone laughs at an early adopter.
Because just maybe, the joke is on those who assumed the old system was the safer bet.
But it never was.
It was just the familiar one.