The Three Fates of Every Skill

The last lesson ended by saying that your skill is not one thing.

That is the single most useful correction available to anyone worried about this subject.

"Am I going to be replaced?" is an unanswerable question, because nobody is employed to be a job title.

People are paid to perform a bundle of tasks, and the tasks in any bundle are moving in different directions at different speeds.

A nurse's bundle contains drug administration, patient observation, family conversations, and three hours of charting.

Those four tasks have entirely different futures, and the nurse's income over the next decade depends on which ones the employer still needs a nurse for.

So the question becomes answerable once you ask it correctly: for each task I perform, which of three fates is arriving?

Every task is heading toward being automated, amplified, or moated.

This lesson defines the three, shows how to tell them apart, and gives you a triage method you will use for the rest of your working life.

Fate One: Automated

A task is being automated when the output can be produced acceptably by a system, at high volume, at a cost that keeps falling.

The signal is not that the machine does it perfectly. It is that the machine does it well enough for the buyer's purpose at a price the human cannot match.

Examples already well advanced: first-draft copy, transcription, routine translation, standard image production, boilerplate code, data cleanup, meeting summaries, bookkeeping reconciliation, first-line chat support, document formatting.

What happens to price is what Lesson 3 described: it falls toward the cost of compute plus a thin margin.

What happens to the people is more varied: some move up a layer to verification and judgment, some move sideways into moated work, and some leave the field.

The important point is timing: you want to know a task is being automated while you still have income from it, not afterwards.

Fate Two: Amplified

A task is being amplified when a person plus a system produces far more than the person alone, and the person is still necessary.

This sounds like the safe fate. It is the most misunderstood one.

Amplification raises output per worker, which means the same demand is met by fewer workers.

Maya's marketing team went from nine people to four while producing more. The four are amplified. The other five were not protected by the same amplification.

So amplified work is good for the individual who survives the consolidation and bad for the population of people doing that work.

And even for the survivor, the gain is only kept if the survivor owns something: the tool, the customer relationship, the contract, or equity.

Amplification without ownership makes you a more efficient employee, not a wealthier one.

That sentence is the bridge between this Part and the rest of the course.

Fate Three: Moated

A task is moated when something other than capability prevents substitution.

There are five common moats, and it is worth being precise about each.

Liability: someone must carry legal or professional responsibility for the outcome. A signature that carries consequences cannot be provided by a system that cannot be sued or struck off.

Physical presence: the task needs a body in a place, with hands, at a specific time.

Trust: the buyer must believe the result, and belief attaches to a person or institution that has something to lose.

License and regulation: entry is restricted by law or by procurement rules that specify a qualified human.

Relationship and access: you hold a connection, a permission, or a position between parties that a competitor cannot simply acquire.

Note what is absent from that list: difficulty, training length, and intelligence.

None of those are moats, because none of them prevent substitution once the capability is abundant.

Moats vary in strength and durability

A moat is not binary. Grade yours.

A license backed by statute and enforced by an insurer is strong and slow to change. A professional norm that "clients prefer a human" is weak and can evaporate in a year.

Some moats are rented rather than owned: your employer's license, your firm's insurance, your platform's account. If the moat belongs to someone else, so does the income it protects.

Most Skills Split

The reason this framework matters is that a single role almost always contains all three fates.

Take the nurse. Physical care and patient assessment are moated by presence and liability, charting and documentation are being automated, and coordination across a shift is being amplified.

Take a lawyer. First-draft contracts and document review are being automated, research and case preparation are being amplified, and appearing in court or signing off on risk is moated by liability and trust.

Take an accountant. Reconciliation and classification are being automated, modelling and scenario work are being amplified, and signing an audit opinion is moated by license and liability.

Take a software developer. Boilerplate is being automated, architecture and debugging across a large system are being amplified, and owning production reliability and the accountability for an outage is moated.

In every case the same pattern appears: the production layer compresses, the judgment layer amplifies, and the accountability layer holds.

Your income in ten years depends on which layer you occupy and, more importantly, whether you own anything in it.

The Triage Method

Here is the method. It takes about an hour and you should repeat it every year.

Step one: list tasks, not job titles. Write out everything you actually did in the last month, in units a stranger could understand. Aim for ten to twenty lines.

Step two: for each task, estimate what share of your income it supports. Rough percentages are fine, but they must add to 100.

Step three: ask four questions of each task.

Who bears the liability if this is wrong, and can that party be a piece of software?

Does this require physical presence at a specific place and time?

Does the buyer need to trust a person, or only to receive an output?

Can the output be verified cheaply, and is a small error rate tolerable?

Step four: assign a fate. Cheap verification plus tolerable errors plus no liability plus no presence means automated, while human accountability plus expensive verification means moated; in between, with a system producing and you directing, means amplified.

Step five: total your income by fate. This is the number that matters.

If more than half of your income comes from tasks in the automated column, you are running a business whose main product is being commoditised, and the clock is already running.

What to Do With Each Column

The triage is only useful if it changes behaviour, so here is the direction of travel for each column.

For automated tasks: stop investing in getting better at them, adopt the tools immediately so you keep the speed advantage while it lasts, and begin moving that share of income elsewhere.

For amplified tasks: do them, and simultaneously ask who owns the upside. Negotiate for ownership of the workflow, the client relationship, a profit share, or equity. If none of those are available, you are building an asset for someone else.

For moated tasks: deepen them deliberately, and check whether the moat is yours or rented. Converting a rented moat into an owned one (your own license, your own client contracts, your own insurance, your own entity) is one of the highest-return moves available to a mid-career professional.

None of this requires quitting anything this month. It requires knowing the composition of your income.

What the Three Readers Do

Maya

Maya's triage produces an uncomfortable split.

Campaign production, copy variants, asset resizing, performance reporting, and briefing documents are roughly 45 percent of what she is paid for, and all of it is automated or nearly so.

Budget accountability, the relationship with the sales organisation, choosing which bets to make, and defending them to an executive are the other 55 percent, and those are moated by trust and by the fact that someone must own the number.

Her strategic error would be to keep proving her value through production speed, because that is the compressing half. Her move is to expand the accountable half and to make herself the owner of the automated pipeline rather than its fastest user.

Tom

Tom's triage is blunt and clarifying.

Straight translation of standard documentation, roughly 70 percent of his historical income, is automated. His rate per word has already told him so.

Terminology decisions on unusual equipment, catching errors that would cause a regulatory rejection, and being the named reviewer who signs off are moated by liability and trust, and they were maybe 10 percent of his old work.

The whole of Tom's plan for the next two years is to invert those percentages, which is a different project from becoming a faster translator.

Leo

Leo's triage looks reassuring and is not.

Answering tickets, writing follow-up emails, preparing onboarding documents, and summarising account health are amplified at best and automated at worst.

His moated column is almost empty, which is normal at 24 and is the thing to fix deliberately rather than by accident.

The two things he can build fastest are relationship depth with a specific set of customers and real knowledge of one industry's failure modes, because both are inputs to the moats he will need later.

Worksheet

This is the triage, done once, properly. Keep the result; Lesson 9 and Lesson 34 both use it.

1. List ten to twenty tasks you actually performed in the last month, described so a stranger could understand each one.
2. Assign each task a share of your income, in percentages that total 100.
3. For each task, answer: who bears the liability if it is wrong?
4. For each task, answer: is physical presence required, and does the buyer need to trust a person rather than receive an output?
5. For each task, answer: can the output be verified cheaply, and is a small error rate acceptable to the buyer?
6. Label each task automated, amplified, or moated, and write one sentence of evidence for the label.
7. Total the percentage of your income in each of the three columns and write the three numbers on one line.
8. For every moated task, mark whether the moat is owned by you or rented from an employer, platform, or firm.
9. Pick the one moated task with the largest growth potential and write the first concrete step to increase its share of your income.
10. Diary a date twelve months from now to repeat this exercise.

Common Mistakes

Triaging the job title instead of the tasks

"Accountant" has no fate. Reconciliation, modelling, and signing an opinion have three different fates.

Every useful conclusion in this lesson comes from the decomposition, so skipping it produces nothing.

Treating amplified as safe

Amplification is the most comfortable and most deceptive column.

It feels like winning because you are more productive and still employed, while the number of people needed for that work is falling underneath you.

Confusing a rented moat with an owned one

Many professionals are protected by their employer's license, contracts, and insurance rather than their own.

That protection is real and it is not yours, which becomes obvious on the day the employment ends.

Assuming difficulty is protection

Work can be extremely hard and still be automated, because difficulty for a human and difficulty for a system are unrelated quantities.

Chess is hard. Folding laundry is easy. Machines took the first long before the second.

Doing the triage once and filing it

Fates move. A task that is amplified this year can be automated in three years, and a moat can be redefined by a regulator.

An annual review takes an hour and is the cheapest insurance in this course.

Fleeing to a field you know nothing about

The most common panic response is to abandon a domain entirely for one that sounds safer.

Your domain knowledge is an input to your future moats, and starting from zero elsewhere usually lands you in that field's most automated layer.

The RW Finance Perspective

This triage is the same analysis we teach for businesses, pointed at a person.

When RW Finance assesses a company, we separate its revenue into lines and ask which lines have pricing power and which are exposed to competition, because a blended average hides the story.

A business with 60 percent of revenue in a commoditising product and 40 percent in a protected one is not an average business. It is two businesses, one dying and one worth owning.

Your income is the same structure, and the triage is how you see it.

We also insist on evidence over narrative. A company that claims a moat but shows falling margins is telling you something, and so are your own invoices.

The Screener exists so investors can filter for quality and durability rather than stories, and this lesson's worksheet is a screen run on your own position.

One more parallel matters. In investing, understanding a business is the step before deciding what it is worth. In this course, understanding your own composition of income is the step before deciding how to convert it.

The next lesson, Skill, Income, Capital, Wealth, supplies the vocabulary for that conversion, and explains why people who are excellent at the first two often never reach the second two.

Key Takeaways

• Nobody is employed to be a job title; people are paid for bundles of tasks with different futures.
• Every task is heading toward one of three fates: automated, amplified, or moated.
• Automation arrives when output is cheaply verified, tolerant of small errors, and free of liability and physical presence.
• Amplification consolidates work into fewer hands, so it protects the survivor and not the occupation.
• Amplification without ownership of the tool, the contract, or the client makes you a more efficient employee rather than a wealthier one.
• Moats come from liability, physical presence, trust, license, and relationships, and never from difficulty or training length.
• Most roles split across all three fates, with production compressing, judgment amplifying, and accountability holding.
• A rented moat belongs to your employer or platform, and converting it into an owned one is a high-return mid-career move.
• The triage is only useful if you total your income by column, because that number tells you how much time you have.