From zero to AI master.
30 days. 5 minutes a day.

Yes. Zekai is an independent AI tools directory — the challenge is how we introduce people to the tools we test. Some tool pages contain affiliate links; the challenge itself costs nothing and requires no purchase.

No, and on the parts that matter it structurally cannot. A model has never stood on your slab, seen your ground conditions, watched your crew work or felt a scaffold move. It can read a contract, draft a method statement and triage a clash list — all of which are real time savings — but every one of those outputs is a hypothesis about a physical place it has never visited. The verification is the job, and the verification requires a competent person on site. What changes is how much of your evening goes on paperwork instead of on the walk.

Installed on day 6, and every other drill in the month is an application of it. Nothing a model tells you about a site is evidence about that site — it has never been there. Every output about physical work is a hypothesis to be verified on the ground by someone competent to verify it, and the closer that output gets to a person’s safety, the earlier and more thoroughly it has to be checked. The numbers behind it are sobering: 1,034 workplace deaths in private-sector construction in the United States in 2024, 389 of them falls, and fall protection has been OSHA’s most cited standard for fifteen consecutive years. This is not a knowledge problem. It is a verification problem.

You can use it to draft the structure and to find the questions you have not asked. You cannot use it to produce the document. A risk assessment and a method statement are things a competent person prepares, signs and stands behind for a specific site, specific plant and a specific crew, and a model has none of that information. Day 7 is deliberately built the other way round: the useful output is not the draft, it is the exhaustive list of everything about your site that only someone who has been there can supply. Nothing in this challenge is a risk assessment, a method statement, or a substitute for competent advice.

That is normal and a ban will not stop it — it will only stop you hearing about it. People reach for whatever answers the question in front of them, and on site that question is often about a detail, a spec or a method. The useful response is day 6: be clear about what it is fine for, what must never be acted on without physical verification by a competent person, and who counts as competent for each category. Publish that this week, while nothing is going wrong, and put it on the wall of the site office.

Because binge-reading builds nothing. The challenge unlocks one drill per day from the day you sign — the way any skill gets built: daily reps. Day 1 is open right now so you can try it before committing.

Zero. Day 1 assumes you have never written a deliberate prompt. By day 6 you will have a written site rule you could pin up in the site office.

Life happens, and so does an unplanned stop. The challenge is sequential, not strict-daily. Pick up where you left off; the certificate cares that you finish all 30, not that you were perfect.

Every edition is built the same way, for a different desk — you can see all of them on the challenge hub, and new professions are added regularly. If yours is not there yet, sign up and we’ll tell you the day it launches.