Harvard Business Review called this back in 2018, before most companies had touched a single AI tool: leadership was about to split in two. AI would take the "hard" half. The "soft" half would become the whole job. Most leaders still haven't reallocated their time to match.
Every AI leadership conversation I sit in eventually arrives at the same anxious question: what does a leader even do once AI can process the data, model the scenarios, and recommend the decision? It's the right question. Most people answer it wrong — because they assume the job just shrinks. It doesn't shrink. It splits.
Tomas Chamorro-Premuzic, Michael Wade, and Jennifer Jordan made this case in Harvard Business Review nearly a decade ago, and it reads more true today than when they wrote it. Leadership, they argued, has always had two components bolted together: a "hard" half — gathering facts, processing information, weighing trade-offs, making the call — and a "soft" half — the personality traits, behaviors, and relational work that get other people to actually commit to that call and act on it.
AI is exceptional at the hard half. It will out-process, out-model, and out-analyze any human leader on raw cognitive throughput, and it already does in narrow domains. It is worthless at the soft half. It cannot build trust. It cannot read a room that's gone quiet in a way that means something. It cannot make a skeptical VP feel heard instead of managed. That work was always human. Now it's about to become the entire job description.
Every leader already accepts, in the abstract, that AI changes work. Fewer have done the specific math on what it does to their own role. If half of what you were hired and promoted to do — synthesizing data, running the numbers, producing the analysis — is about to be done faster and more consistently by a tool, then roughly half of your historical value proposition as a leader is being commoditized in real time.
That's not a threat. It's a reallocation problem, and it's exactly the kind of problem PROSCI's ADKAR model was built to solve — except this time the "change" isn't a new system rollout, it's the leader's own job. I've written before about what a manager's job becomes in the AI era; this is the leadership-level version of the same shift, one rung up the org chart.
Skip any one of those five and you get a leader who keeps doing the hard-half work out of habit, competing with a tool that's faster and cheaper, while the soft-half work — the part no one else in the org can do — goes unattended.
This is where I see the most expensive blind spot in organizations right now. Companies are hiring AI strategy help to fix the technology side — the models, the data pipelines, the governance frameworks. Fewer are hiring anyone to help their leadership bench figure out what their job becomes once the hard half is automated. That's not a technology consulting problem. It's a change management and leadership-development problem, and it's the exact seam an AI management consultant is built to work.
The organizations that get this right aren't the ones with the best AI tools. They're the ones whose leaders deliberately moved their time and attention toward the half of the job AI can't touch — before the market forced the reallocation on them.
Audit the split. Sit down with your leadership team and map, honestly, what percentage of each leader's week is hard-half work (that AI already does or soon will) versus soft-half work (that only a human can do). Most leaders have never done this exercise and are surprised by the ratio.
Reinvest the time, don't just cut it. When AI takes three hours a week off someone's plate, that time needs to be redirected toward the soft half — coaching a struggling report, walking the floor, having the hard conversation that's been deferred — not absorbed into more meetings about AI.
Build the soft-skill muscle deliberately. Trust-building and judgment under ambiguity aren't personality traits you either have or don't. They're trainable, coachable, and measurable — the same way financial acumen is. Treat them that way.
Change what gets rewarded. If your performance reviews still primarily credit "drove the analysis," you're reinforcing exactly the half of the job that's being automated. Start explicitly crediting the alignment-building, resistance-navigation, and trust work that makes any analysis actually land.
AI didn't shrink leadership. It clarified it — by taking the half that was always more measurable and easier to hire for, and leaving behind the half that was always harder to teach and more valuable to have. The leaders who see that clearly, and reallocate accordingly, are the ones whose teams will still be following them in three years.
The AI Efficiency Audit maps exactly where your organization's hard-half work is already being automated — and whether your leaders have reallocated toward the half that matters.
Start with an audit →Sources: Chamorro-Premuzic, T., Wade, M., & Jordan, J. (2018, January 22). As AI makes more decisions, the nature of leadership will change. Harvard Business Review. hbr.org | MIT xPRO. (2024). AI Strategy and Leadership Program, Module 9: Data Analytics Is Decision Driven. Massachusetts Institute of Technology.