Everyone Wants AI. Nobody Wants to Do the Work That Makes AI Valuable.

One of the most interesting things about the current AI conversation is how little we talk about the work that happens before AI.

Most organizations are discussing copilots, assistants, prompt libraries, use cases, and productivity gains. Vendors demonstrate impressive capabilities, leadership teams ask how quickly AI can be adopted, and employees are encouraged to experiment with new tools.

Yet very few conversations focus on what determines whether any of those initiatives will actually succeed.

In my experience, the biggest challenge isn't selecting the right AI platform. It's whether the organization has built the foundations that AI depends on.

That work is rarely visible.

It happens when HR teams spend months standardizing job titles across countries, defining job families, cleaning workforce data, agreeing on skills definitions, documenting processes, reviewing governance, or deciding which system should be considered the source of truth. These projects don't usually receive much attention because they aren't particularly exciting. They don't produce immediate results, and they often compete with initiatives that seem more innovative.

Ironically, they're exactly what makes innovation possible.

I've noticed that AI has changed the way many organizations think about these initiatives. Work that was once viewed as operational suddenly becomes strategic because AI depends on consistency. If two similar roles have completely different titles, AI has no way of understanding that they're related. If the same skill is defined differently across business units, recommendations become inconsistent. If workforce data exists in multiple systems without clear ownership, AI won't resolve those differences. It will simply use whatever information it can find.

That's why I don't think AI creates organizational maturity.

It reveals it.

Organizations with disciplined People Operations, clear governance, reliable workforce data, and a well-maintained job architecture usually see value from AI much faster. The technology isn't necessarily better. The environment it's operating in is.

The opposite is also true.

When AI produces inconsistent recommendations, inaccurate summaries, or unreliable insights, our first instinct is often to question the model. Sometimes we should. But just as often, AI is exposing issues that have existed for years. Incomplete data, fragmented processes, inconsistent definitions, and unclear ownership were already creating problems. AI simply removes many of the manual workarounds people had developed over time.

That's why I think the conversation around AI readiness needs to become broader.

Readiness isn't measured by the number of licenses you've purchased or the percentage of employees who have completed AI training. It's reflected in the quality of the foundations underneath. Can people trust the data? Are roles defined consistently? Is there governance around skills, jobs, and workforce information? Are processes standardized enough that AI can support them instead of introducing more variability?

Those questions aren't as exciting as discussing the latest AI capabilities, but they'll have a much bigger impact on whether organizations see meaningful results.

Technology will continue to evolve. Models will become more capable, faster, and more accessible. That's no longer what differentiates organizations.

The differentiator is becoming much more familiar.

Strong People Operations.

Reliable workforce data.

Thoughtful job architecture.

Clear governance.

Consistent processes.

None of these initiatives became important because AI arrived. They've always been important. AI simply made it impossible to ignore them.

Perhaps that's one of the most valuable things AI has done for HR.

It reminded us that the quality of our outputs will always depend on the quality of our foundations.