For students: build the capacity, not the shadow

Planned for release phase 2. The page is complete as a draft; interactive parts run in the browser and do not yet save to an account.

Why the system broke, and why it is not your fault

In college, the work product is not the point. Your assignments are not the point; your grades are not the point. The point is the building of the underlying capacity, and the grade is only a shadow of it. In a normal year the shadow and the capacity track each other. The moment the shadow can be produced without the capacity — the moment a tool can write the assignment — the system breaks, and it breaks silently, because the grade keeps coming. The rational thing to do, if the grade is the only signal you are watching, is to optimise for the shadow. Most students are. That is not laziness. It is a broken signal.

1. Tutor, not substitute

There are two questions you can ask any AI tool. Do this for me. Teach me how to do this. The first gets you an output and the output gets you a grade. The second gets you the shape of the thinking, which you then do yourself. Ask it to explain. Ask it to walk through the reasoning. Ask follow-up questions until you actually understand. Do not ask for the answer. This is the whole difference: not whether you use AI in college, but whether you use it to skip the climb or to make the climb faster. The first hollows you out. The second compresses years of learning into months.

2. A portfolio you can defend

Your professor cannot tell, in most courses, who built understanding and who outsourced it; the institution's signal has decayed. So build your own signal, on your own time, that you can trust. Spend the equivalent of one course's effort every semester on a project of your own choosing. A small game. A useful tool for your hostel. A data analysis of something you actually care about. It does not have to be big. It has to be yours — every line of it something you could walk a stranger through. By graduation, four or five of these. In 2026 and after, that portfolio is worth more than your transcript, because the transcript can be faked and the portfolio cannot.

Run each project through the framework. Step 1 for a student is not "finish the assignment"; it is "what do I want to be able to do without help when this is done?"

3. One assignment a week, no AI

Pen, paper, brain. Slower than your classmates. Worse output than you would produce with the tools. That is fine; the point is not the output. The point is to keep the muscle alive and to know, every week, that you can still do it. The students who will be in trouble in five years are the ones who never had this practice — who graduate with grades that do not reflect capacity, and freeze in their first job when a senior person asks them to walk through a problem, and do not know why they froze, because the freezing is not about that problem. It is about four years of outsourced ones.

The students who do this will have both: the capacity, kept alive, and the speed, from using the tools when speed matters. They will be — the book's word — thick, where most people are about to become thin.

The thing you will not believe

This is the best possible time to be where you are. Not the worst. You are scared, and you think the world is about to leave you behind. It isn't. You have years before you have to be somebody specific. The people in trouble are the ones in their thirties who built a career on capacities being automated and have to reinvent in real time with mortgages and children. You get to do the reinvention before you have built anything. That is an enormous advantage. You won't believe it. It is true anyway.