Earlier this month I joined the People Is Not an Option podcast for their Decoding Talent special edition, and we spent the better part of an hour pulling at a thread that I think the field is still underestimating: the reason AI transformations stall in HR organisations is rarely about the technology. It is almost always about culture. The podcast is hosted by Alessandro Rimassa and Simone Patera.
As the Podcast is in Italian, here a summary of my key answers. My starting point was something I have observed across multiple transformations. Traditional HR technology was, for most of its history, a late arrival — built to record what had already happened, to manage numbers and administrative transactions. AI is structurally different because it works with language: it reads context, pulls inferences, recognises patterns in everyday communication. That single shift has profound consequences for what HR professionals actually do, and for the skills they need to do it well.
The discomfort I see in the profession right now is real and worth naming honestly. AI is targeting precisely the relational, experience-based expertise that HR has traditionally positioned as its core value. That is uncomfortable. But I'd rather treat it as a signal than as a threat — a signal that the pivot required is not technical upskilling but something closer to what I'd call philosophical literacy: the ability to frame a context, articulate a problem well, ask the right question. When I was training early conversational tools, the thing that determined quality of output was not engineering knowledge. It was the ability to think clearly and express intent precisely. Philosophy courses, it turns out, were more useful preparation than SAP certification.
We talked through the Campari experience at some length — specifically the work we are doing on commercial capability building. The honest lesson there is that AI is an extraordinary tool for mapping what already exists inside a permeable workforce before you decide what to build or buy. The old model — three years of competency mapping resulting in a 100-page booklet that was obsolete the day it was printed — is not a problem of effort. It is a problem of rhythm. AI allows you to move iteratively, to focus on two or three priorities at a time rather than trying to codify everything at once.
I also flagged two failure modes I see repeatedly. The first is distraction: AI generates so much possibility that it is easy to drift away from the actual business problem you started with. The second is postponement: waiting for the "right" corporate tool while employees are already using consumer-grade AI on their phones, outside any governance structure. Both mistakes are understandable. Neither is acceptable.
The question I was most interested in — what does HR need to own in this environment — has a straightforward answer that the profession is still resisting: the redesign of work itself. Not job administration. Not position management. The foundational questions: what does work look like when AI handles an increasing portion of the cognitive load? And why would someone choose to do that work here rather than somewhere else?
Comments and Feedbacks
More posts like this
