Minds for Our Minds at Work

Executive Summary

In 2016, we found that the work most resistant to automation was the work full of unpredictability—messy people, unknown environments, situations that change while you're in them. We're happy that it held for a decade. But then generative AI started handling unpredictable situations on its own, and unpredictability turned out to be a symptom, not the cause.

We think that the cause was really irreducibility. Some work can be pulled apart into pieces, solved piece by piece, and reassembled. Some cannot, because each part is conditioned by all the others at the same time—pull them apart and the thing you were doing stops existing. That is the actual line between what AI absorbs and what stays human.

So we've built an index to measure it. The Irreducible Complexity Index scores 894 US occupations on five reasons work won't come apart: whether it needs a body in a place, whether someone must own a contested call, how many knowledge domains it braids together at once, whether it requires solving problems nobody has posed yet, and whether it demands a sense of taste no rubric can capture. Each is a different kind of resistance, and progress on one does nothing to resolve the others.

The most resilient occupations combine several of these at once. Most work, including most knowledge work, relies on only one or two. That concentration is the vulnerability so broadening it is the opportunity.

When we placed AI's current capability on the same scale as human skill requirements, the conventional career advice inverted. The skills we're told to chase—programming, mathematics, analysis—are where AI is furthest ahead. The interpersonal and physical skills are where humans lead widest. What we used to call soft skills we should perhaps now call higher skills.

The replacement story—the machine does what you did, and you go find something else—is wrong in a way the scores alone don't show. We have spent years studying how people actually work with AI, and what we keep finding is people doing more, not less. The routine layer gets absorbed and the harder work underneath surfaces. Developers who used to write syntax now put product, domain, cost, and user together and become builders. We've seen analysts who used to format data now making the call the data was meant to inform. The circle of what work asks of people doesn't shrink, it expands.

That is what we call authorship, and it points to what we think AI should actually be for. Not efficiency or replacement but extension and expansion. Minds for our minds at work—tools that let a person reach further into the irreducible work, balance more of a system at once, make hard decisions with better support. In our fieldwork, we consistently observe that the people who flourish with AI are the ones who argue with it, direct it, and use its output inside their own judgment. At this point, we assert the difference is awareness, agency, and accountability—though we treat this as an observation that warrants formal study, not a demonstrated causal claim.

The IRX makes testable predictions. It predicts that occupations concentrated on one or two dimensions will see faster displacement than those spread across several, and that AI will continue to absorb separable subtasks from high-scoring occupations while leaving the integrated core intact. If AI begins performing whole occupations that score above 70—not pieces of them, the entire integrated practice—the framework is wrong. Validating it against employment and displacement data as it accumulates is required.

This has practical application. If resilience comes from combining multiple kinds of irreducible difficulty, you need to build real capability across dimensions you're thin on. Add physical or operational grounding to cognitive work. Build genuine cross-domain knowledge. Develop aesthetic judgment and a deep intuition for what is better. That advice is easier to follow with time, resources, and security, and the transition will not be equally accessible—structural support matters alongside individual adaptation. But the work ahead is not shrinking, it is growing, and minds for our minds is how we think it should be built.