Glossary

The fixed-mindset trap that catches senior leaders with AI: the more your identity is tied to knowing your domain, the more your practice quietly avoids the work where it would matter most.

Three sequential moves that pull a leader back from the AI-expanded option space to a single committed page: close the diagnosis, commit a guiding policy, name coherent actions.

Three components, Corpus, Curriculum, Contract, that turn the proxy a leader is already training accidentally into one that represents them by design.

Leaders who reject AI and leaders who claim it as their own are practising the same failure: a provenance judgement made before a merit judgement.

Six layers, model, data, training signal, values, override, exit, across which a leader retains accountable ownership of the AI proxy operating in their voice.

Six cognitive biases that bend AI use: three before the leader prompts, three after AI replies.

The organising axis of the AI for Leaders series, running from identity-absent (workslop) to identity-total (identic AI), along which every AI-assisted output is already set somewhere.

Three named recursions, output, process, practice, that turn one AI session into compounding thinking and prevent single-pass plateau.

AI removes the friction that built leaders’ judgement; the discipline is to put it back, by design, in three named places.

Three probe questions that intervene between an AI’s contradiction and the leader’s reflex, turning disagreement into information rather than threat or verdict.

The discipline of asking whether a different framing of the question would make the obvious answer unnecessary.

Three foundational management mechanisms have shifted under generative AI, leaving the old reasoning correct and the conclusions wrong.

The discipline of getting 1% better at one specific dimension of your AI use each working week, embedded before the next, so capability compounds rather than plateauing.

The cumulative cost of skimming AI output instead of working it through; fourteen named costs across four quadrants, paid in quality now and capability later.

Every AI session opens in one of two operating modes; the one chosen in the first thirty seconds shapes the output more than any prompt that follows.

Three moves, challenge, check, capture, that turn an AI hour from maintenance into an investment that compounds session by session.

A 2×2 that plots every AI session by cognitive demand and time-allocation choice, and shows that AI’s strategic leverage lives in only one of four named quadrants.

Three components, context-aware, accountable, durable, that distinguish the leader’s call from the model’s recommendation when the assistant can argue both sides equally well.

A leader’s stance towards AI as a domain where capability is built at the edge of current ability, not protected by the appearance of already knowing.

The deliberate willingness to be visibly slower and clumsier with AI inside your own domain of expertise, paid in exchange for capability that does not arrive without that cost.

Four structural elements to redesign and three to preserve, to close the gap between desk-level AI productivity and firm-level results.

Routing a task upward to AI rather than downward to a team member, when handing it down would erode self-determination and the task is mechanical or contextually low-value to the receiver’s role.

The senior-leader misread that treats an AI productivity problem as a motivation problem, when more than eighty per cent of the gap in most organisations is structural.

Three mechanical questions that move a leader from AI dashboards to constraint dashboards, so AI deployment lifts organisational throughput rather than producing only motion.

The quiet moment-by-moment decision to stop engaging critically with AI’s output, taken when engagement feels expensive and acceptance feels free.

Leaders default to using AI for the easiest, lowest-leverage tasks because friction is near-zero and the reward is immediate.

Five named relationships a leader can take with AI, each fitting different work, each chosen on purpose before the prompt.

AI is a worker. Brief it, choose it, calibrate its autonomy, check in, and own what comes out.