We've been putting AI to work inside real companies since ChatGPT landed in November 2022. In that time the thing that surprised me most had nothing to do with what the technology could do. It was how differently people took to it. For some of us, myself included, working with AI came quickly and kept getting easier. For others, capable and motivated people who wanted it to work, it stayed stubbornly hard. Same tools, same training, same encouragement. Very different results.
That observation could have stayed a curiosity. It became something more because of Brett Hurt, the co-founder and CEO of the company I worked at before this, data.world, who kept pushing me on a point I was slow to absorb. Doing impressive things with AI was the easy part. The harder and more valuable work was the change management around it, how people actually fold a new capability into the way they already work. The interesting problem was never the model. It was the person on the other side of it.
Every client since has reinforced it. The pattern shows up across industries, roles, and levels of technical comfort, and it keeps looking the same. A company will tell us their AI adoption is poor and expect us to know what that means. It almost never does on its own. One company means their people barely open the tools. Another has teams doing advanced work and still feels behind. Both walk in with the identical sentence. We can't help either of them until we know which one we're looking at.
That is the trouble with "adoption." It sounds like a measurement, but it collapses a dozen very different situations into one word. You can hand every employee a license, run the training, watch usage climb, and still have changed almost nothing about how the work gets produced. Usage is not the same as people producing better work in a genuinely different way, which is what leaders actually want.
So we use a more precise term for the distance that actually matters: the AI Assimilation Gap. It is the space between using AI and meaningfully incorporating it into how you produce work. Someone who opens ChatGPT every morning to answer a few questions and someone who has rebuilt their workflow around AI are both, technically, adopters. The gap is everything sitting between them, and it is where most of the people we work with live.
We didn't invent the idea from nothing. Most people picture technology adoption through Everett Rogers' diffusion curve, the familiar spread from innovators and early adopters out to the majority and the laggards. That curve sorts people into types. What we keep running into is a different distance, one that sits inside a single person or team, the space between having a technology and actually putting it to work. Robert Fichman and Chris Kemerer named the organizational version of that distance in 1999, calling it the assimilation gap and showing how a company could acquire a technology long before it had deployed it in any real sense. Grant, who leads this thinking at Gadoci Consulting, has taken their term to the personal level, mapping the individual version, the one you feel in your own work with AI. The framework is his.
The gap exists for reasons that are more human than technical. Some of it is belief. You carry conclusions about yourself and your work, like "I'm not a programmer" or "I need to produce this myself to stand behind it," that stay true in your head long after the conditions underneath them have changed. The rest is cost. Changing how you produce work you already do well means paying real time and patience up front, before you know whether the new way is better, while the old way keeps finishing today's work without complaint. Neither makes someone resistant or behind. Both are the ordinary price of changing how competent work gets made, and putting names to them is where we go next.
Naming the gap is where the useful conversation starts. Once a company can see the difference between "our people don't use AI" and "our people use AI but haven't changed how they work," the right kind of help finally comes into focus. That focus is most of what we do.