Does Behavioral Psychology Hold the Key to Unlocking the Power of Synthetics in Research?
Synthetics hold promise for research, but without human-like drivers they often fail to feel real or trustworthy. By combining behavioral psychology with a layered generational taxonomy, we can model not just responses but the deeper decision-making processes behind them. This integration points toward a future where synthetics become more reliable digital twins, better aligned with how people truly think and behave.
I’ve spent the last few years deep in the world of AI — building systems, experimenting with synthetics, and testing the limits of digital twins. Some of the work has been exciting. Digital twins tied to real, known people can be surprisingly reliable and useful. They mirror human behavior closely enough to provide insights we can work with.
But pure synthetics? That’s a different story.
On the surface, they function. They answer questions, they generate data, and they look like they’re working. But something is missing. They don’t feel human. The cracks start to show in simple places, like Likert scales. Real people bring biases, habits, and inconsistencies into their responses — the “messy humanity” that makes research data so challenging and so valuable. Synthetics, on the other hand, often default to balanced distributions or neat patterns. And that isn’t how people behave.
Why Pure Synthetics Struggle
The core issue is that pure synthetics lack drivers.
Humans don’t make decisions in a vacuum. Every choice is filtered through experiences, biases, and beliefs. We skip questions, lean on heuristics, contradict ourselves, and sometimes change our minds halfway through a survey. Those quirks aren’t noise — they’re the very texture of human behavior.
Synthetics can be forced into better alignment through training, weighting, or constraints. But those fixes feel artificial — because they are. What’s missing is the deeper set of drivers that guide people long before they pick up a survey or walk into a store.
What My Generational Research Taught Me
In researching Between Silicon and Soul, I built a layered taxonomy of the generations. It wasn’t just a list of stereotypes or activities. It was a framework that explained why people behave the way they do.
It drilled below the surface of consumer actions — streaming, job-hopping, brand choices — into the deeper cultural, generational, and psychological forces that shape them. That taxonomy helped me understand my own kids and their friends, nieces and nephews, and a rising generation that sees the world through a lens my Gen X peers often struggle to grasp.
That work taught me an important lesson: surface-level behaviors are only the tip of the iceberg. If you want to predict or model people, you need to understand the drivers underneath.
Where Behavioral Psychology Comes In
This is where behavioral psychology becomes so critical.
Behavioral science gives us frameworks for how people actually make choices: cognitive biases, heuristics, habit loops, loss aversion, social proof. These aren’t just academic theories — they’re patterns of decision-making that we see everywhere in real life.
Now imagine combining those behavioral insights with a layered generational taxonomy. The taxonomy explains who someone is and what motivates them. Behavioral psychology explains how those motivations show up in their decisions. Together, they form a foundation for synthetics that don’t just mimic answers, but actually behave like people.
A Modern Bayesian Approach
The way I see it, this is essentially a modern Bayesian model for synthetics.
Real people don’t generate responses randomly, and they aren’t purely rational calculators. Instead, we all operate with priors — beliefs, values, and experiences — and we update those priors as we encounter new information.
That’s what makes people so fascinating. Given who I am, what I’ve lived, and what I’ve seen, how likely am I to respond this way?
If synthetics can be structured to operate on those priors — drawing on both cultural/ generational drivers and psychological patterns — they stop looking like simulations and start looking like something closer to digital twins of real human behavior.
Why This Matters for Research
At the end of the day, the core issue with synthetics in research is trust.
If we can’t trust them to act like people, they remain clever toys. But if we can ground them in real drivers of behavior, then synthetics can become reliable tools for foresight, testing, and innovation. They can allow us to experiment with the future in ways that traditional methods can’t.
This isn’t about replacing people. It’s about building more faithful models of how people think, feel, and decide — so that research keeps pace with the changing world around us.
The Bigger Realization
As I’ve been working in this space, I’ve come to a bigger realization: the whole of me — father, researcher, technologist, writer, maker, believer — is more useful than the compartmentalized version of me.
When I integrate my work in AI, my taxonomy of generations, my observations as a dad of Gen Z kids, and my passion for understanding the human spirit, the insights are deeper. The solutions are better. And the work is far more meaningful.
Maybe that’s the real lesson here. It’s not just synthetics that need integration. It’s us.
Closing Thought
So does behavioral psychology hold the key to unlocking the power of synthetics? I believe it does. Especially when paired with frameworks like my layered taxonomy.
It’s not perfect yet. Validation is still hard. But this direction feels right — because it’s bringing us closer to modeling people as they really are, not as neat simulations of how we wish they behaved.
And maybe that’s the next frontier: not just making machines act like humans, but learning how to integrate our own knowledge, disciplines, and humanity into something greater. Between silicon and soul.