This guide assumes three things: you are not going to stop AI, you cannot predict the timeline, and you would rather act than worry. Work through it in order. It takes about an hour to read and about a quarter to execute.
Step 1 — Score your exposure honestly
Exposure is not about your job title. It is about the tasks inside your job. Write down everything you did last week and mark each item:
- Red — a model can already do this at 80% of your quality, at 1% of the cost. Text production, first-draft analysis, routine code, basic design, tier-one support, standard research.
- Amber — a model helps a lot but you are still needed. Judgement calls with context, client relationships, complex debugging, editorial decisions.
- Green — requires a body, a licence, accountability, or trust that has to be earned in person.
If more than half your week is red, you do not have a career problem yet — you have a two-year window. Use it.
Step 2 — Run the readiness check
// Readiness check
How adapted are you right now?
Step 3 — Move up the value chain
There are four positions in an AI economy. The goal is to move rightward.
Task doer
You execute the work personally. Directly substitutable, paid by the hour, and the hour is getting cheaper.
Tool operator
You get more done because you use AI well. Better, but the tools keep getting easier, so the advantage decays.
System designer
You build the workflow that produces the work, and it runs whether you are there or not. This is the durable position for most people.
Owner
You own the asset, the audience, the catalogue or the equity that the systems produce value for. This is where the upside of automation actually lands.
Step 4 — Diversify your income structurally
One employer paying you for tasks is a single point of failure. Aim for three legs, of different types:
- A cash-flow leg — employment, consulting, contract work. Stability now.
- An asset leg — a product, catalogue, app, book or audience that earns without your hours. Start one this quarter: print on demand, books, apps.
- A human leg — work that pays because it is you: teaching, advising, community, presence. The most defensible thing you own.
Step 5 — Build the skills that keep paying
- Problem framing. Turning a vague situation into a specified task is the skill AI makes more valuable, not less.
- Verification. Knowing when an output is wrong, fast. This requires domain depth — do not abandon your expertise, weaponise it.
- System design. Decomposing work into steps, tools and checkpoints. See forty workflows for worked examples.
- Distribution. Making things is now cheap; being found is not. This is the real bottleneck for everyone building an asset leg.
- Judgement under uncertainty. Covered properly in singularity or doom.
Step 6 — Protect the basics
Unglamorous, and it is what actually determines whether a disruption is a setback or a catastrophe: reduce fixed costs, build a cash buffer measured in months, keep your professional network warm enough to call, and keep your health — the one asset with no substitute.
What not to do
Do not quit a stable job to chase an AI gold rush with no runway. Do not put your savings into a single token or a single stock because a video told you to. Do not assume your field is uniquely protected. And do not spend four hours a day reading AI news — that is consumption wearing the costume of preparation.
Next
Turn this into weeks: the 30-day action plans. Pick your tools: the directory. Understand the game: how to play AI.