How to Be Future-Proof

Three letters, multiplied

Being hard to replace is three things at once: one skill somebody can name and buy (T), one industry you know from the inside (B), and a standing habit of noticing what is coming (F).

They multiply. Five shallow skills do not add up to one deep one — they multiply toward zero. A brilliant specialist in an industry they never studied is a hammer looking for a nail, and gets priced like a hammer.

The ladder: what goes first

Work sorts into five rungs. Repeatable digital work with no judgement is already gone. First drafts and standard outputs are going now. Rule-bound analysis has a few years. Judgement under ambiguity is slow to automate. Accountability — someone must sign, someone must be blamed, someone must be believed — is the safest thing there is.

It is not about how hard the task is. It is about how ambiguous it is and who carries the consequence when it goes wrong.

Score any task in ten seconds

Is it repeatable? Is it fully digital? Is it judgement-free? Is it consequence-free? Four yeses and you are standing on the bottom rung. Four noes and you are underpriced.

Do this for the ten things you actually did last week, weight each by the share of your week it took, and add up the red. That percentage is your exposure.

Being good was not protection

The uncomfortable finding from the research: after the models arrived, top-rated freelancers and average ones were hit the same way. Quality did not show up as protection in the data.

Being excellent at a commodity task was never a strategy. It had simply never been tested before.

Four things that get MORE valuable

Judgement — deciding what is worth doing, including what to refuse. Trust — being the person who can be held responsible. Taste — knowing which one of the hundred is right for this audience. Agency — starting before anyone asks you to.

And the fifth nobody teaches: selling. The other four are worth nothing if the person writing the cheque has never heard of them.

The radar costs ninety minutes a quarter

Three signals. Use the newest tool in your field yourself — not read about it — and ask what it learned to do this year. Follow the capital, because money moves twelve to twenty-four months before jobs do. Read twenty job ads at your level and twenty one level above.

The warning: 'I tried it and it was bad' is exactly what every displaced expert said. Judge the trend, not the version.

Three levels of using AI

Most people ask questions and copy answers — a smarter search engine, worth maybe a tenth more output. Building repeatable workflows that do your recurring work is where the premium starts. Owning systems that run without you, and selling the output instead of your hours, is where the money is.

The trap is using it to do more of the thing being automated. Getting faster at work that is losing its price is running harder toward the cliff.

Skills expire, assets compound

A portfolio, an audience, a network and a stake all survive the tool that kills the skill. A skill lives inside somebody else's process and stops paying the moment you stop working.

The only question worth answering: if the work disappeared tomorrow, what would you still own?

Your note

Saved on this device as you type.