Wunmi Alimi
Managing Director · 6 min read
Board pressure to "do something on AI" tends to convert very quickly into a headcount request, and headcount requests tend to convert into a job description scraped from whatever similar-sounding titles are trending on LinkedIn that quarter. The role gets a title, Head of AI, AI Transformation Lead, Director of AI Strategy, well before anyone has defined what decisions it actually owns.
The hire arrives, and finds no operating model to plug into, no governance forum that recognizes their mandate, and a set of stakeholders who all assumed the new hire would solve a different problem than the one any of them actually had in mind. Within a year, the best outcome is that they quietly become a generalist technologist doing adjacent work. The more common outcome is that they leave, and the organization concludes, wrongly, that AI talent is impossible to retain.
The sequencing problem is the root cause, and it's fixable. Define the target operating model first. Derive the specific capability gaps and accountable outcomes that model requires. Only then write the role brief, grounded in what the person needs to actually be accountable for, not in what a market scan suggests other companies are calling the role.
This sequencing also opens up options that get skipped when hiring runs on autopilot. Sometimes the right answer to a capability gap is a permanent senior hire. Often it's a fractional executive who can establish the function while the permanent role and its mandate get properly defined through real operating experience, rather than guessed at in advance.
It also changes how you should evaluate candidates once the brief is right. Delivery evidence, a portfolio of shipped work, demonstrated judgment in genuinely ambiguous situations, predicts success far better than a list of certifications or the prestige of a previous employer's AI lab.
None of this requires a better job board or a more aggressive recruiter. It requires sequencing: capability model before role, role before requisition. Most of what gets called an "AI talent shortage" is, on inspection, a capability-definition shortage wearing a recruiting costume.
“Most "AI talent shortages" are capability-definition shortages wearing a recruiting costume.”