Two ladders: why "I can do this" and "I earn from this" are different things
45.2% use neural networks regularly. 91.9% have earned nothing or earned once. That is not a contradiction — these are two different ladders, and the same person stands on different rungs of each. Below: how it came about, how the rungs differ, and what stops working once you understand it.
45.2% use neural networks regularly. 91.9% have earned nothing or earned once. That is not a contradiction — these are two different ladders, and the same person stands on different rungs of each. Below: how it came about, how the rungs differ, and what stops working once you understand it.
---
The figure that does not add up
I run a Telegram channel with no announcements for the platform — purely out of interest in the subject. A year ago I did one live session and that was that. In August I decided to ask my readers directly what they do and what gets in the way — 62 replies.
The picture was not the one I expected.
45.2% answered "I use it regularly". Daily or nearly. They pick models to suit a task, pay for subscriptions, build complicated things.
54.8% have not earned a single dollar. Another 37.1% only earned once. AI has not become the main income for anybody.
At first glance that is a contradiction: people can do it and do not earn.
There is no contradiction. One respondent stated the diagnosis in five words: the gap between doing and monetising.
---
Two ladders
There are two independent paths, and a person moves along them at different speeds.
| | The tool ladder | The money ladder |
|---|---|---|
| Where most people stand | rung four or five: can do it, picks models, pays for subscriptions | rung two or three: knows people earn from this, does not know how |
| What they need | depth, access, no ceiling | a price list, a place to find work, a script, a first case study |
| What is useless | "look what a neural network can do" | "here are twenty more services" |
You are almost certainly high on the first and low on the second. That is the normal position for most of this market.
---
Why it turned out this way
Three reasons, and none of them is about your ability.
People teach the first ladder because it is easier
Showing how a tool works is easy: share a screen, press a button, get a result. The material takes an hour to make.
Explaining how to name a price, where to find a client and what to say to "that's expensive" is work of a different order. And it requires sales experience, not button-pressing experience.
The figure that bears this out. I went through 6 502 posts from eleven large channels in this niche over 19 months. Posts containing price guidance: nine. Nought point one per cent of the feed.
It is not that the subject is uninteresting. It simply requires something most authors do not have.
The first ladder is infinite
Models come out every week. There is always something new you have not tried.
That creates the feeling that you are not ready. The feeling is false, but it is convenient: while you are not ready, you do not have to enter the market.
There is no feedback
On the first ladder you are your own judge: it came out nicely, so it is good. On the second the judge is the client, and they either pay or they do not.
The second is more honest and therefore more frightening.
---
What stops working once you have understood this
The practical part, because understanding ought to change what you do.
"One more course on a tool" stops working. If you already use AI regularly, another course on neural networks will move you from rung four of the first ladder to rung four and a half. On the second you will not move at all.
Reading AI news stops working. It addresses a rung you have already passed. A post about a new model lands in a void not because the subject is bad but because it settles nothing for you.
"I'll finish learning first" stops working. The moment of having finished never arrives: the industry supplies a fresh reason to continue every week.
And something else starts working: a price, a place to find work, an application, a first case study. Dull things, none of them about technology.
---
What the second ladder looks like
Five rungs. Find yours.
One: never thought about it. I use it for myself, I have not tried to earn. What you need: to see that people pay for this at all, and how much.
Two: I understand it is possible but not how. What you need: a route. One service → one tool → one place to find work → a first payment.
Three: I am trying and it is not working. Applications go out, no replies. What you need: to work out where exactly it breaks — the price, the packaging, or the application itself.
Four: one-off jobs exist, a flow does not. What you need: to move from units to a service and learn to hold a client for a second job.
Five: the flow exists. What you need: to raise your rate and take production off your own hands.
The biggest drop is between rungs two and three, and that is where people fall most often. Because that is where you have to stop preparing and start sending.
---
How your height on the first ladder does help
The good news, which gets lost behind the diagnosis.
You do not need to learn how to make things. You already can, and that is a large part of the path plenty of people have not walked.
You need to learn four things, and none of them takes months:
Naming a price — that is a table of ranges, ten minutes to read.
Finding where the work is posted — that is one place, not fifteen.
Writing an application — that is a template adjusted for a particular client.
Showing your work — ten pieces in one niche, assembled in an evening.
Three evenings in total. Not three months.
The months go elsewhere: on sampling tools and preparing to prepare.
---
What to do today
One action, ten minutes.
Mark your position on both ladders. Honestly, with no rounding in your own favour.
If you are on rung four or five of the first and rung two or three of the second, you are standing where most of this market stands. Nobody writes about it: it is awkward both for the people who did not earn and for the people who taught them.
And the second thing. Look at what the last piece of AI material you read or watched was about. If it was the first ladder, you spent your time on a rung you had already passed.
---
What is ready for this on the platform
The ten-lesson sales module — the whole second ladder: where to find clients, four client types, handling objections, dealing with rejection, selling without case studies, shaping an offer around demand. Included in the Basic plan.
"Freelancer positioning" — the four things listed above in one pass. Seven assistants, and what comes out is documents: a niche, a client avatar, a portfolio, profile texts, a plan for your first jobs. $9, included in Basic.
The job feed — a daily sweep of Telegram chats, filtering by a model, cards with the client's contact. 29 unique requests a day.
Registration is free and opens three days of full Basic access. Three days is exactly the time it takes to work through all four points.
---
What to read next
[AI freelancer price list](/en/blog/ai-freelancer-price-list) — the first of the four points: nine types of work and the range for each.
[Where to find AI work](/en/blog/where-to-find-ai-orders) — the second point: 29 requests a day and how to see them.
[A portfolio without a single client](/en/blog/portfolio-without-clients) — the fourth point: ten pieces in an evening.
[Five reasons a skill does not turn into money](/en/blog/skill-but-no-money) — each barrier taken apart in the words of the people themselves.
[AI API pricing in 2026](/en/blog/llm-api-pricing) — what a request costs and how to work out your own costs.