The Distribution
87 percent of people are followers. NPCs by design. They show up to the script someone else wrote and play the part as written.
10 percent are destroyers. They have leadership capacity and they use it badly. They lead movements, companies, congregations, and crews toward outcomes that diminish everyone they touch. Capable, charismatic, off-mission. The 10 are awake and pointed in the wrong direction.
3 percent are kingdom leaders. They lead for the perpetuation of good. They build institutions, practices, and rooms that outlast them. They use power to widen the circle of people who can use power well.
These numbers are approximate and the spirit is exact. Every community I have been close to maps onto them. The destroyer/kingdom split is the leadership-tier version of the two teams.
The Proportions Hold
Self-improving AI does not move the distribution. The 87/10/3 is a fact about what people are formed to do. Give an NPC a self-improving Jarvis and you get a high-functioning NPC. Give a destroyer one and you get a more efficient destroyer. Tooling is downstream of formation.
Anyone selling AI as the great equalizer is selling a version of the story that does not survive contact with how human beings actually allocate their time.
The Ceiling Moves
AI raises the ceiling. A kingdom leader with a Personal Agentic OS and a real PRM operates at a multiple no individual leader could touch in 2015. They run more institutions, hold more relationships, refactor more systems, draft more strategy, and stay calm while doing it.
The good news is the 3 percent can do dramatically more. The honest news is the 10 percent can do dramatically more too. The middle does not move.
The Shift That Is Actually Available
The interesting question is whether the 10 can be pulled toward the 3.
The 10 are morally flexible, capable, and primarily optimizing for status and money. They lead in whatever direction those incentives point. If the incentives point toward kingdom outcomes, the 10 produce kingdom outcomes. They are recruitable on terms that respect their actual operating function.
Two things move them:
It makes them look better. When the cool people are on the kingdom side, the morally flexible operators show up. Reputation is a real currency. See make sovereignty cool: the same mechanism applies to every kingdom posture. The 10 will pay almost any price to be in the room the 3 are already in, as long as that room is the room everyone else wishes they were in.
It is more profitable. When the kingdom side has the better deals, the better partners, and the better long-run multiples, the 10 show up. See aligned self-interest: selfish people backing the right work is exactly how durable institutions get built. Make kingdom outcomes the 100x choice and the 10 walk through the door themselves.
The Job
The job of the 3 is to build institutions, products, and rooms attractive enough to pull the 10 in. Make the kingdom side cool. Make it profitable. Wire the incentives so a morally flexible operator looking around at age 45 says: "the smart move is to back what they are building."
Every Belafonte recruited a hundred Brandos. Every Stowe gave a hundred politicians cover. The 3 percent needs to be effective enough that the 10 want to be seen with it, and recruitable enough that the deal both sides actually want is on the table.
Self-improving AI makes this easier on the 3 side, because impact-per-leader compounds. The math underneath stays the same. Most people will not lead. Some will lead badly. A few will lead well. The play is to make the few impossible to ignore and easy to join.
The distribution stays. The ceiling rises. The 3 lead better, the 10 destroy better, the 87 keep being the 87. The leverage is in pulling the 10 into the 3 by making the 3 the room everyone wants in.