Copilot to autopilot
A copilot suggests, drafts and assists beside the human. An autopilot takes the controls on whole tasks. The line between them is drawn by judgement.
Sequoia’s framing of where AI is heading has two stages. Copilot: the model sits beside the human, suggesting, drafting, assisting, and the human stays on every task. Autopilot: the model takes the controls on whole tasks and reports back. Most account teams are in the first stage and assume the second is a technology question, about when the models get good enough.
It is not. The line between the two is drawn, revealingly, by judgement. Procedural work, the research pack, the meeting summary, the first draft, moves to autopilot early and safely. Judgement-heavy work, the proposition, the negotiation, the read on whether a sponsor is wavering, stays with the human for as long as being wrong is expensive. The question for a KAM is not whether to hand over the controls but which tasks, and the five ages are the longer version of the same answer.
Set out in full in: Sequoia Capital, via Investec / RIT Capital Partners; Creating Customer Value Propositions with AI (Kogan Page, 2027), chapter 2.
Related terms
Human in the loop
Adopt, but keep your hands on the wheel. The machine amplifies judgement. It does not supply it.
The five ages of AI
Prompt, agent, collaborator, advisor, architect. Where you are, where the work is going, and which end is still informed speculation.
AI agent
Software that takes a goal, plans the steps, uses tools and keeps working until it is done. On the buyer's side it is the procurement of the future; on yours it maintains the plan.
See the method run on one real account in three days: the One Page Proposition.