Early in my career, I worked as a market research analyst on the vendor side. I shared a cubicle with four coworkers, and once a week, we read horoscopes together. Our fearless leader, a Sagittarius, ran the readings. She pulled them from Betches.com (a very reputable astrology source) and added her own commentary as she went. Sometimes a Capricorn from the next cubicle over would pop her head over into our cubicle to hear hers. This was mostly a bonding experience, a way to connect and take a break. But every once in a while, a horoscope would feel weirdly accurate. It would match weekend plans, how we were feeling about work, or truthfully, where we were at with our dating life.

That meaning didn't come from nowhere. Astrology, of course, is built on constellations. And constellations have been around for about 5,000 years — Babylonian astronomers first drew them. The Greeks renamed them after their heroes. The Romans translated them into Latin. By the time my coworker was reading the Sagittarius horoscope off Betches.com, humans had been finding meaning in those same stars for millennia.

Whether or not you believe in the stars, the interesting part isn't the horoscope. It's how we found meaning from it.

At its core, this is what researchers do. We make meaning out of patterns — in the sky, in data, in our work, in the small things we notice about each other. Anthropologist Clifford Geertz called it "webs of significance." The meanings we spin around the world, the patterns we draw between things, the stories we use to make sense of what we see.

But in a corporate context, meaning has a deadline. It doesn't get to arrive slowly. There are clients to please, stakeholders to keep on track, decks to build, and decisions to be made.

We've all heard it: AI integration isn't optional anymore. Researchers are feeling that pressure too. According to Qualtrics' 2026 Market Research Trends Report, 39% of research leaders say AI has revolutionized their workflows. Yet only 19% of individual contributors agree. I'm right there with you.

In fact, just a few weeks ago I was working with some trended data. I figured I'd let ChatGPT do the first pass. It came back quickly with a finding: two groups were diverging in their views. At first glance, a clear pattern.

But when I looked closely at the data myself, I wasn't sure the pattern was strong enough to report. The numbers moved, but not by much. Sure, AI detected a pattern emerging, but my perspective on the data was that it was too early to assign meaning.

There's a difference between a pattern and a finding. AI can tell you something moved, flag statistical significance, and summarize hundreds of qualitative transcripts in seconds. What it can't tell you is whether any of that means something — to this client, in this category, for this organization, given everything else you know about the research. That part requires something AI isn't built for: slowing down.

So what does this all mean for researchers? The reality is, the thing AI can't quite do is the part that needs time. And if you ask me (my actual hot take) AI will never be able to do this part meaningfully. Geertz would agree. Meaning-making, at its core, is inherently human. It requires cultural context, lived experience, and interpretation that AI can pattern-match but never actually inhabit.

The strongest argument for AI in research is speed — faster, more efficient, more productive — and I understand the need for that efficiency. I've lived it myself, stuck in the endless research process, completing one monotonous task after another. It can be easy to fall into this space, where the data and insights tell you something, or they tell you nothing at all.

This is the part that keeps me up at night. And I mean that literally, some of the best thinking I do happens when I'm sitting beside my toddler's crib, waiting for her to fall asleep. Meaning, for me, almost always arrives somewhere other than my desk.

There's another idea from cultural anthropology that echoes this one. Daniel Miller, an anthropologist who has spent his career arguing that the meaningful stuff of culture doesn't live in grand rituals — it lives in ordinary objects and small daily routines.

The same is true of research. Meaning comes from ordinary moments. It can be on a walk in between meetings, getting your morning coffee, or catching up with a coworker. And I'll admit, sometimes it's even in an AI chatbot.

Still. We have to slow down. That's when the good ideas come through. This doesn't mean we should abandon AI, but rather be thoughtful in how we use it and take a step back every once in a while. When I'm working with AI, I often step away from the conversation and return to it the next day. Sometimes what happens next is hard to explain, almost mystical. The aha moment arrives. A sense of how to move forward that wasn't there before, and sometimes you just can’t explain where you got it.

This essay, in and of itself, is evidence of Daniel Miller's idea in action: meaning lives in the ordinary. I came up with the idea for this piece beside my daughter’s crib, singing Twinkle Twinkle Little Star to her.

The patterns are everywhere — in my daughter's lullabies, in the stars, in the data, in our work. The meaning is what we make of them. AI can find the patterns. But we still decide what becomes a constellation.