Part III · Chapter 7 · Leverage · lesson 2

Which kinds need permission

10 min

Builds on Why one hour of some work is worth a thousand of another

To hire someone, a person has to accept the job. To borrow, a lender has to approve the loan. To publish, nobody has to do anything at all.

That is the whole distinction, and it sorts the six multipliers into two groups that behave completely differently over a life.

Some leverage requires a specific person to say yes before it exists. Some does not.

Permissioned leverage is faster to obtain and can be withdrawn. Permissionless leverage is slower to build, pays nothing for a long time, and once it exists nobody can take it back. Neither is superior. They fail in opposite directions, and knowing which kind you are relying on tells you what to worry about.

Sorting the six

Needs permission: labour and capital.

Labour needs an employee who agrees, and money to pay them before the output arrives. Capital needs a lender who agrees, and lenders agree on the basis of what you already have — which is the mechanism behind the observation that borrowing is cheapest for people who need it least.

Both can be arranged inside a week by somebody with a good position, and both can be removed. A team resigns. A lender declines to renew a facility that has rolled over uneventfully for six years. Neither event requires you to have done anything wrong; the permission was always somebody else's to give.

Needs no permission: technology, media and knowledge.

Writing code requires a machine. Publishing requires an account. Learning requires attention and time. None of the three involves an approval, an interview or a credit assessment, and none of them can be revoked once the thing exists.

The price of this is that nobody pays you to start. There is no first paycheque for building an audience, and the first two years of any of the three produce output that reaches almost nobody. Permissionless leverage is free to acquire and expensive to wait for.

Somewhere in between: network.

A network is built out of many small yeses given over years, no one of which was a gate. Nobody approves your network, and it cannot exist without other people's willingness, so it behaves like a permissionless asset built out of permissioned parts.

Something people say

“You need money to make money.”

Why it spreads: It is visibly true of the multiplier people can see. Capital is the leverage that shows up in the story of every large fortune, and it genuinely is the fastest one, so a rule built only from those stories comes out this shape.
The part that is true: Starting capital is an enormous advantage and this course will not pretend otherwise. It removes the need to be paid quickly, which is the constraint that stops most people using the slow multipliers at all.

What the saying misses is that three of the six multipliers cost nothing to acquire and are not available for purchase at any price. Nobody can buy a decade of learning, an audience, or code they have not written. What money actually buys is the ability to wait while the slow ones compound — which is a real and large advantage, and a different claim from the one the saying makes.

The trap in the middle: rented distribution

There is a category that looks permissionless and is not, and it catches careful people.

Publishing on a platform requires no approval, so it reads as permissionless. But the audience is reached through a system whose rules can change, whose ranking can change, and whose relationship with you is terminable without explanation. What you built is real. The channel it travels down is rented.

The distinction that survives is between an audience you can reach directly — a list of people who gave you an address, a product they installed, a relationship they would maintain elsewhere — and an audience you can only reach while an intermediary keeps forwarding you. Both are worth having. Only one of them is yours.

Predict

Somebody has built a following of two hundred thousand on a single platform over four years, and earns well from it. In what sense is this permissioned leverage?

It is not — nobody approved the account, and the work is entirely theirsThe reach depends on a ranking system that can change without noticeBecause the earnings come from advertisers rather than from the audienceIt is permissioned only if the platform charges a fee

Not the same thing

Permissioned leverage

A multiplier that exists because a specific party agreed, and stops if they stop agreeing.

Measured in: How quickly it can be obtained, and how quickly it can be withdrawn
For example: A credit facility; a team of employees; a distribution agreement.
Chase it alone and: It is available exactly when you least need it and withdrawn exactly when you most do, because the party granting it is reading the same conditions you are.

Permissionless leverage

A multiplier you can create unilaterally, which nobody can revoke once it exists.

Measured in: Years of unpaid input before it produces anything
For example: A body of published work; a working system you built; what you know.
Chase it alone and: It pays nothing for long enough that most people stop, and there is no external signal telling you whether stopping was sensible or a mistake.

Which is which? Put each one on a side.

An overdraft facility that has been renewed every year for a decade
A tool you built that automates four hours of weekly work
Ten thousand people who have given you a direct way to contact them
Top placement in a marketplace that sends most of your customers

What this implies about sequencing

Neither category is the right answer, and the two interact in a way worth naming.

Permissioned leverage is obtained on the strength of what you already have, so it is easiest to get after the permissionless kinds have produced something. Permissionless leverage requires a long unpaid stretch, which is easiest to survive if something else is paying the bills.

That is not a paradox, it is a sequence, and it is why the slow multipliers are usually built alongside a job rather than instead of one. It is also why the advice to quit and go all in is so much more dangerous than it sounds: it removes the thing funding the wait, right at the point where the wait is longest.

Check

Two people both stop working for a year. One spends it building an audience; the other spends it managing a leveraged position funded by a loan. Both fail to make money in that year. Whose position is worse at the end of it?

The audience builder — a year of income given up for nothingThe leveraged one, because the loan is still owed and grewThe same — both lost a yearNeither — a year is too short to matter

Worth remembering

Labour and capital require a specific party to agree, and can be withdrawn by that party.
Technology, media and knowledge require no approval and cannot be revoked once they exist.
Permissionless leverage costs nothing to acquire and pays nothing for a long time, which is its real price.
Distribution on a platform is permissionless to create and permissioned to reach, so the channel is rented even when the work is owned.
Permissioned leverage is granted on the strength of what you already hold, which is why it tends to arrive last.