Most public conversation about AI and leadership stops at the obvious surface — productivity, automation, cost reduction. That is the boring version of the story.
The interesting version is what a small set of senior operators are quietly doing with AI behind closed doors. They are not optimising their inbox. They are systematically increasing their leverage over the three things that actually decide the next ten years of their career: how they are seen, who they get to work with, and what their work produces.
I will call those, in order: narrative, talent, and outcomes. They are the three levers. AI changes how you operate each one.
Lever one: narrative
Narrative is who you are in the rooms you’re not in.
Senior leaders tend to underweight this for one of two reasons. Either they were taught that the work should speak for itself (it doesn’t), or they associate narrative work with a flavour of LinkedIn theatre they find embarrassing (which is fair, but a category error). The compounders treat narrative as critical infrastructure and engineer it deliberately.
Three changes AI has made to narrative work:
The cost of producing high-signal public writing has collapsed. Not because AI writes for you — that would be a different thing, and visibly worse — but because the workflow around your own thinking is now substantially faster. Outline, draft, sharpen, second-pass, version, test, ship. A single piece that used to take eight hours of attention takes two if you’re working with a strong model as a sparring partner.
The frequency floor has moved. It used to be that publishing once a quarter was enough to keep your name in the right rooms. The compounders now publish once or twice a month, at higher quality. The gap between occasionally seen and consistently seen has widened.
Positioning can be engineered explicitly. What I mean: the way you want to be seen — what you’re known for, what you stand against, what you’re building toward — can be drafted, tested and refined like a product. Not by performing a personality. By naming, in writing, what you actually believe and what you intend to do. AI makes the iteration loop on that fast.
The compounders almost universally have a deliberate, written, periodically-revised positioning. The administrators almost universally do not.
Lever two: talent
Talent is the second lever and the most under-discussed one.
Senior leaders tend to think of talent as something HR does. The compounders treat it as their primary multiplier. AI changes how they operate it in three specific ways.
Sourcing is no longer a calendar problem. Identifying the strongest operators in a given function used to take meaningful research time. It now takes one structured prompt, a public-data search, and a thirty-minute review. The compounders maintain a live, AI-assisted talent map — not because they are recruiting today, but because the right hire shows up six months before you knew you needed them.
Reading people is faster. When the compounders need to evaluate a candidate, a board member, a counterparty, they read everything that person has written publicly in an hour, with AI as a research partner. They walk into the conversation already understanding the person’s priors, blind spots and signature moves. The administrator walks in with the candidate’s LinkedIn and a stale referral.
Retention is run as a continuous process, not an event. The compounders maintain ongoing structured notes on the people who matter most to their work, with AI helping them stay specific. They remember what people care about, what they’re working through, what they’re reading. The strongest operators below them notice. People stay where they are seen.
The pattern that ties all three together: AI does not replace judgement about people. It removes the friction that used to prevent senior leaders from acting on their judgement at the right moment.
Lever three: outcomes
Outcomes is the lever that looks the most boring and is the most decisive.
Almost everything about senior work is a sequence of decisions: which deal, which hire, which strategy, which trade-off, which fight to have, which fight to skip. The quality of the decisions, taken together, determines the outcome of the year.
The compounders use AI to operate decisions differently in three ways.
They write the decision before they make it. A short, structured document — what the decision is, the options, the considerations, the recommendation, the risks. Drafting it forces clarity. AI accelerates the drafting and improves the analysis. The administrator decides verbally in a meeting and rationalises it later.
They run pre-mortems on the consequential ones. Before committing, they ask the model to generate the most plausible failure modes — and then a few they hadn’t considered. This is unflashy and disproportionately useful. The administrator is told about the failure mode by someone in week six.
They keep a running journal of decisions and their actual outcomes. This is the move that most differentiates the most senior people I work with from everyone else, AI-assisted or not. They look back, regularly, at what they decided and what happened, and they update their priors. AI makes the journal cheap to maintain and the review cheap to perform. The administrator doesn’t track and doesn’t calibrate.
What this looks like inside a week
A composite week of a senior compounder, drawn from people I work with:
- Monday morning. A 90-minute session with a model: what landed last week, what didn’t, what to compound, what to drop. Output: a one-page weekly directive.
- Across the week. Memos, reviews, strategy documents drafted with the model as sparring partner. Edited heavily. Final version is the operator’s — but they reach the final version several times faster, and several drafts deeper, than they used to.
- Two short pieces of public writing. Either a LinkedIn post or a private newsletter to a vetted list. Engineered against the deliberate positioning. Written in twenty minutes apiece, not two hours.
- One structured decision document. For the most consequential decision of the week. Pre-mortem included.
- One quiet talent move. A note to a person whose work they are tracking. A reach-out to someone they have been meaning to bring closer. A consequential nudge to someone on their team.
- Friday afternoon. Twenty-minute review with the model: what compounded this week, what didn’t, what to change next week.
That is the entire game. There is no special tool, no exotic prompt, no secret model. There is a small set of habits, run with discipline, with AI as the partner that makes the habits affordable.
The thing nobody is selling you
You don’t need to become a content creator. You don’t need to become a programmer. You don’t need to know which model is best at which task this month. You need to make AI part of how you operate the three levers — narrative, talent, outcomes — for long enough that the difference is visible.
The compounders aren’t doing anything mysterious. They are doing the obvious things deliberately, and they are doing them every week.
If this is the kind of work you want to do alongside a small group of senior operators making the same shift, the Strategic Influence Program is built around exactly this material.
If your situation is more particular and you would prefer to work on it privately, the 1:1 Executive Mentorship covers it directly.