A working philosophy
AI changes the shape of work.
The useful question is not whether your company has adopted AI. It is what changes when research, drafting, analysis, and coordination become cheaper—and how you redesign ownership, structure, and judgment around that fact.
Operations
It’s not about the tools. It’s about what they change about operations.
A model entering a workflow changes more than cycle time. It changes how:
- decisions are framed and pressure-tested before a meeting;
- knowledge moves from conversation into reusable operating context;
- work is delegated, reviewed, and returned without losing intent;
- teams coordinate when drafting and synthesis no longer set the pace;
- managers allocate scarce attention once preparation gets cheaper;
- accountability is drawn when output is abundant but judgment is not.
That is operating redesign, not a software rollout or a training exercise. Tool proficiency may help, but it cannot decide which meeting should disappear, where a handoff should change, or who remains answerable for the result.
Capacity and structure
Shifts in capacity come with shifting structure. Shift accordingly.
Time saved is not yet capacity gained. It becomes useful only when someone decides what disappears, what changes hands, or what newly deserves attention. Without that decision, cheaper preparation can produce more documents, more review, and the same bottleneck.
Stop
Remove work whose purpose was to compensate for slow retrieval, manual synthesis, or information that arrived too late. Do not preserve a report or meeting merely because it is familiar.
Reassign
Move preparation and first-pass analysis to the role closest to the evidence, with explicit review boundaries. The transfer should clarify ownership rather than create a new executive queue.
Reinvest
Direct recovered attention toward customers, talent, strategy, and work the organization has deferred because urgent preparation kept winning.
The structural question follows the capacity question: if the work changes, should the role, handoff, meeting, or decision right change with it?
Judgment
Additional context need not cloud judgment. The decision is still yours.
A model can widen the field of view, make assumptions visible, and connect a decision to evidence that would otherwise remain scattered. None of that transfers the obligation to decide.
Use AI to
- surface alternatives;
- retrieve context;
- challenge assumptions;
- trace evidence.
Keep human-held
- the stakes;
- evidentiary judgment;
- tradeoffs;
- consequences.
The model expands available context. The executive determines what counts, resolves the tradeoffs, and owns what follows.
The wager
The wager: consider the null hypothesis.
AI is not the answer must remain a live possibility. A serious operating review tests whether changing the workflow would improve anything that matters, not whether a model can be made to participate.
Leave a workflow alone when:
- There is no material constraint.The work already moves at the speed and quality the business needs.
- Context loss would outweigh speed.The tacit knowledge, relationship, or interpretive nuance is the work rather than friction around it.
- Error or consequence cost dominates the benefit.A faster first pass is not useful when a plausible mistake is difficult to detect, reverse, or absorb.
Leave it alone can be the right operating conclusion.
Practice
Hands-on, or it stays an abstraction.
AI may work in theory. The test is what survives actual context, evidence, constraints, and accountability.
Illustrative executive working examples—not client case studies or promised outcomes
Pressure-test a board narrative
- Operating moment
- The board materials are factually complete, but the argument may hide assumptions or leave predictable questions unanswered.
- Useful AI contribution
- Test the narrative against alternative readings, identify claims that need evidence, and expose tension between the stated outlook and supporting detail.
- Human decision and ownership
- The executive decides what the board needs to understand, what uncertainty to disclose, and which position the company will defend.
Find the variance worth discussing
- Operating moment
- A forecast or operating review contains more movement than the meeting can examine well.
- Useful AI contribution
- Compare current results with plan and prior assumptions, then surface variances, pattern breaks, and questions that merit verification.
- Human decision and ownership
- The leader validates the evidence, chooses what deserves airtime, and determines the response.
Compare organizational structure before moving people
- Operating moment
- A proposed reorganization solves one visible bottleneck but may shift decision rights or create new handoffs elsewhere.
- Useful AI contribution
- Lay out structural alternatives, map role and dependency changes, and challenge the assumptions behind each option.
- Human decision and ownership
- The executive weighs capability, trust, careers, and business consequences before changing anyone’s role.
Turn scattered context into a decision brief
- Operating moment
- Market, regulatory, and customer signals sit across notes, source documents, and separate conversations.
- Useful AI contribution
- Organize the evidence, distinguish agreement from conflict, trace claims to sources, and frame the open decisions.
- Human decision and ownership
- The executive judges source quality, decides what the combined context means, and selects the action.
A working conversation
Bring the operating question that will not resolve itself.
The Operating Leverage Session is a paid 50-minute working session with a principal. Bring one current AI question and leave with a recommendation for the next step. The fee is credited in full toward any engagement.
One-on-one work with the executive—not implementation, team training, legal advice, or a guarantee of results.