There has always been a hierarchy. Most years, it shifts at the edges. Once a generation, the ground moves underneath it.
The last time the ground moved at this speed was probably the early days of the internet. The leaders who treated it as a marketing channel optimised the wrong thing for a decade. The leaders who treated it as a new operating layer for capital, attention and talent built the firms we now refer to by their first name.
AI is the next moment of that kind. The reason most senior leaders will miss it is not that they are unaware — every executive committee has a slide about AI by now. The reason is that the meaningful shift is not the slide. It is what happens, quietly, in the rooms below the slide.
What's actually happening
Three things are running in parallel right now.
Knowledge work is being unbundled. The activities that used to require a junior team — research, drafting, comparison, structured reasoning, document review — are increasingly being done by a single operator with a model. Not perfectly, but well enough that the marginal cost of a thousand decisions has collapsed. Senior leaders who already had high judgement now have judgement and throughput. The people without it are quickly becoming an expensive bottleneck.
The narrative layer has become a moat. Every senior person now has a public surface — LinkedIn, conference speaking, podcast guesting, written essays. AI doesn’t make that surface easier; it makes it possible to maintain at a much higher quality and frequency. The people doing it well are pulling away from the people doing it shallowly. There is no middle anymore.
Talent is sorting itself. The strongest operators below you are noticing — quietly — who is using AI to compound their work and who is using it as a faster Google. They are recalibrating who is worth following, who is worth reporting to, and who is going to age out. Most of this happens in private channels you can’t see.
The two layers
If you take a slow look at any senior leadership team right now, you can usually identify two distinct layers.
The compounders. Five percent, maybe ten. They have made AI part of how they think — research, drafting, decisions, communication. Their week looks different from a year ago. They produce more, with cleaner reasoning. They are present on more public surfaces, with higher signal. They are quietly hiring people like themselves and quietly disengaging from people who can’t keep up.
The administrators. Everyone else. They use AI sometimes, in shallow ways, on small tasks. Their week looks roughly the same as it did two years ago, with marginal improvement. They are not getting worse — they are simply being passed.
A leadership team that is 80% administrators and 20% compounders looks fine on a quarterly slide. It feels increasingly tense in the rooms where the compounders meet.
Why most leaders will miss it
Three reasons, in order of how often I see them.
It looks like a tooling decision. It isn’t. A tooling decision happens once, propagates downward and either works or doesn’t. This is an operating-model decision, made daily, in a dozen small choices about what you do yourself, what you delegate, what you read, what you write, who you spend time with, and what you publish. Tooling decisions get made in committee. Operating-model decisions only get made by individuals.
The early gains are quiet. A leader who compounds for twelve months doesn’t produce a flashy result. They produce a slightly different version of themselves. Sharper. Faster. Better positioned. More frequently in the right room. The story is invisible until you compare them to a peer who didn’t. Then the gap is striking and very hard to close.
It feels uncomfortable to discuss directly. Most senior leaders are not comfortable saying out loud, I am behind on this and I need to catch up. So they don’t say it. They schedule another all-hands about the company’s AI strategy and quietly hope the personal version of the question goes away. It doesn’t.
What to do about it
The honest answer is unfashionable. There is no shortcut. The compounders did not arrive at where they are by attending three webinars and buying a Pro subscription. They arrived by deciding, on purpose, to make AI part of how they work — and then doing it for nine to twelve consistent months.
That said, three moves materially shorten the curve.
Replace one habit a week. Pick a recurring activity — your Monday review, your direct-report 1:1 prep, your inbox triage, your strategy doc draft — and run it through Claude or a comparable model for six weeks. Not as an experiment. As your new default. After six weeks, you keep what works, drop what doesn’t, and pick the next one.
Engineer your public surface deliberately. This is the one most senior leaders avoid the longest, and the one with the largest delta between I’m thinking about it and I’ve made it real. The bar isn’t becoming a content creator. The bar is making sure the senior people you want to work with in three years know who you are and what you think.
Audit who you spend time with. The talent layer below you is sorting itself faster than you are sorting it. The strongest operators want to be near other compounders. If your room is mostly administrators, the compounders will quietly leave for someone else’s room. The best leaders I work with are unusually deliberate about this.
The conversation worth having
Most senior leaders I speak with already know they are slightly behind. They don’t know how far, they don’t know what specifically to do, and they don’t want to have the conversation in front of their team. That is the entire reason this practice exists.
The hierarchy is not going to wait for a more comfortable moment.
If this is the conversation you have been putting off, the Strategic Influence Program is built for it.