At work you sit on the side that offers something, and the other side arrives prepared: with an assessment of their own and a price range produced by a model. Your knowledge advantage no longer carries your price. What carries it is your judgment and the responsibility for the outcome, and you have to name both before you start working. So settle early who is writing along on either side, and where the unpaid preliminary conversation ends.
The same person who is strengthened by AI as a private customer in the evening sits on the other side of the table in the morning: as a self-employed person, a coach, a freelancer, a service provider. And there they experience the flip side of their own new power. What strengthens the customer weakens the provider, and the working individual is both at once. That is the uncomfortable symmetry of this chapter.
On the provider side the collapse of the knowledge gradient hits hardest. The self-employed person used to sell exactly what the AI now gives away: knowledge, framing, a first draft. The consequence is already visible in the markets. In freelancing there is a yawning gap between those who bill by effort and those who bill by value: whoever sells the value earns markedly more at the median than whoever sells the hour [Jobbers benchmark, cited in Plutio 2026]. Developers report the oppressive feeling of quoting a price for two hours of AI-assisted work that feels like fraud, even though the result has the same value as before [CodeToDeploy/Medium 2025]. The working individual thus faces the sharpest variant of the double consulting burden from Chapter 3: they must refine the customer's half-knowledge and build their service on top of it, and the first part usually happens unpaid, before the contract.
Jobbers benchmark, cited in Plutio 2026
At the same time, and this is the hopeful half, it is precisely the working individual who is the most agile. A large organization cannot change its business model overnight; a self-employed person can. They can be the first to stop selling hours and start selling judgment, responsibility and outcome, the three things the AI does not deliver. They can make the pre-phase visible and paid because they need ask no one's permission. And they benefit doubly from their double role: because they themselves experience, as a private customer, what it feels like to walk into a conversation AI-armed, they understand their own customer better than anyone who sits on only one side of the table. The working individual is the party that experiences the protocol most directly in hard currency, as protection of their margin and as the language in which they meet their equally half-knowing customers without lecturing them.
Your double role is an advantage once you name it
You know both sides of the table, and that is worth more than any negotiating tactic. You know how certain a machine answer feels, because you have walked into a conversation carrying one yourself. So do not hold it against your customers that they arrive prepared. Ask instead about the assumption their preparation rests on, because that is where the effort sits, and that is where your contribution sits too.
The same holds inside the company. If you are employed, what counts is not only what you deliver but which version of your work arrives further up. If you lead a team, what counts is whether your team knows what it may do with which tool. In both cases the same sentence helps as with a customer: say who or what did the writing here, before anyone has to ask.
Example: the enquiry that arrives as a finished concept
An enquiry came as a short email: what is needed, by when, and a request for a quote. You built the frame in the first conversation, and that conversation was the start of the collaboration.
The enquiry arrives as a finished concept with a schedule and a price range someone had generated overnight. You spend two hours explaining which assumption in it does not hold, and those two hours appear in no quote. If you end up above the range, your price looks like something that needs justifying rather than something that is right.
The enquiry comes from an agent that queries several providers in parallel for the customer and compares the answers before a person reads them. Your explanation therefore reaches a system first, one that sorts by price and delivery time. What sets your proposal apart from the others does not appear in that sorting.
Example: the application and the appraisal
You applied with a cover letter and a CV, and at some point a person in HR read both. Whoever invited you could tell you why.
Your cover letter is written with an assistant, and on the other side a system sorts the incoming applications before anyone sees them. Two machines write past each other. You learn neither which criterion filtered you out nor whom you could ask about it.
Inside the company there is mediation too. Your appraisal draws on a summary of your project reports that nobody on your team wrote. Which version of your work arrives further up therefore belongs in your employment terms rather than to chance.
Example: the draft from the team
Whoever put something forward in the team had built it themselves, and you could ask how it came about. Uncertainty was audible, because people hesitate when they do not know something.
The drafts arrive faster and they all sound equally certain. You can no longer tell from a document whether someone thought it through or merely passed it on. Your team has no idea which use you consider acceptable, so nobody talks about it, and nothing gets checked.
You write down what a tool may do in the team and who is accountable for the result. Whoever writes that down early does not have to guess later where a figure in a document came from.
Draw the line before you start
Say two things at first contact, and say them in writing.
- What you do for free and from which point your work is paid. A look at a concept someone brings along is work, even when it is short.
- What you are accountable for. Name what you check and answer for yourself, and what you do not take on because it comes from a source you do not know.
When the other side arrives with a result a machine produced, ask about the assumption inside it before you talk about the price. The assumption decides the effort, the number in front of it does not.