AI skills beat years of experience. Learn them for free.

Questions

The things people ask before they save a seat.

Every answer is the honest one. Where the answer is no, it says no.

1. Do I need to know how to code?

   No. Nobody writes a line of code in this course. The whole toolchain is a chat app you already have, a spreadsheet, and one free AI app you install on day one.

2. How much does it cost?

   Nothing. The first 1,000 seats are free. There is no card to reserve a seat and no card afterwards. Later cohorts stay free too, with fewer seats each time.

3. Do I need a paid AI subscription?

   No. Every session is built to be finished on the free tier of the apps you already have, and the AI model you install on day one is free too. No API keys, no credit card.

4. Which AI apps does it use?

   The four big ones, ChatGPT, Claude, Gemini and Grok, plus a small open-weight AI model such as Gemma or Qwen, installed free with PocketPal AI on a phone or LM Studio on a computer, so you see what the big products add around a bare model. You meet all four and work hands-on with at least three, and the final comparison runs one task across three of them and your own local model, scored on a sheet you wrote first. Free accounts are enough for all of it: one on day one, and two more free signups later in the course, on the days that need them.

5. Can I really do the whole thing on a phone?

   Yes. Every session, deliverable included. Not a reduced version, the same work, and a phone-only learner earns exactly the same certificate. If your phone cannot run the small local model, you pair with a cohort peer and do the same work from supplied screenshots, for the same certificate. One of the two people who built this course finished a twelve-month software engineering programme on a tablet.

6. How much time does it take?

   Four sessions of 90 minutes a week, which is six hours, plus Day 5 for the weekly quiz and catching up. Four weeks and 16 sessions in total: 24 hours of sessions.

7. Are the sessions live, or do I go at my own pace?

   You run each 90-minute session yourself, and there is no lecture to sit through. What is fixed is the day. This is a cohort rather than a self-paced course, so everyone is on the same session on the same day, and Day 5 of each week is where the week is closed off. Each session is a page you open on your phone.

8. What happens if I fall behind or miss a week?

   Day 5 is the catch-up day and exists for exactly this. Miss two sessions in a week and you use Day 5 and the weekend, with no penalty and nothing recorded. Every quiz has two retakes. After the four weeks there is an open week to rest and catch up before the next course.

9. What do I walk away with?

   Four things. A small AI model running on your own phone or computer that works with the internet off, a notebook of your own prompts with the versions that made them better, a comparison of the big AI apps and your own local model that you ran, scored and can show somebody, and the course certificate.

10. What are the quizzes, and what does it take to pass?

    Ten questions at the end of each session, eight right to pass, two retakes. Forty questions on Day 5 of each week, thirty-two right to pass, two retakes. They are multiple choice with one right answer and you mark them yourself, against an answer key that gives a reason for every answer. For the certificate you need to pass 13 of the 16 daily quizzes and all four weekly ones, and quizzes are not the whole bar: it also needs your project submitted and two peer reviews given.

11. Is the certificate accredited?

    No. We are not an accredited school and this is not a degree. The certificate names what you actually did: the sessions you finished, the quizzes you passed, the project you submitted, and the two peer reviews you gave.

12. Is this a prompt engineering course?

    Yes, and it starts a step earlier than most. Prompt engineering is the craft of asking for something in a way that gets you a usable answer. Week one teaches how the machine works, so the asking is a decision rather than a guess. Week two is the asking itself: the four things missing from most weak requests, why saying the reason beats stating the rule, when examples work better than explanations, and a famous technique you will test yourself and find does not do what it promises. Then you measure whether your new way of asking beat the old one.

13. Is the course finished?

    No. Introduction to AI and LLMs opens on 14 September, and its sessions are being written right now, in the order you will take them. Each one is human-tested before it reaches you. More courses come after this one, shaped by what Cohort 1 teaches us, so their names, order and content can all move. Two things will not move: every course stays free and runs four weeks.

14. What comes after this course?

    More short courses, later this year, on the same terms: free and four weeks. Introduction to AI and LLMs comes first for everyone, because everything else builds on it. After that you pick and choose: take the courses your work needs, skip the ones it does not, and each one earns its own certificate. What we build next follows the needs of the market, our partners, and the people taking the courses.

15. When does it start, and what if I miss the date?

    Cohort 1 starts on Monday 14 September 2026. If you miss it, reserve a seat anyway. Later cohorts stay free, with fewer seats each time.

16. Who runs it?

    Julien Barbier and Youssef Charif Hamidi. Julien co-founded Holberton School and was Chief Product Officer at ALX, where the programmes went from a few thousand learners to more than 1.6 million. Youssef builds the platform, and is the reason we know the phone path works.

17. Can my organisation run this for its people?

    Yes. Cyberiad Labs creates and licenses programmes: run one of ours with your people, or ask us to build one for your needs. The partners page explains how it works and how to reach us.