For decades, building software meant learning to think in English first. The keywords, the documentation, the answers you searched for at 1 a.m., the frameworks, and most of the people teaching them were overwhelmingly Western and English-first. If your first language was Kinyarwanda, Tagalog, or Quechua, the price of entry was never just logic and effort. It was a second language, learned well enough to read a stack trace.
That filter never measured who had a builder's mind. It measured who happened to be born close to the tools. Plenty of people with real technical intuition — the instinct for systems, for edge cases, for “what if we did it this way” — never got to find out, because the on-ramp was written in a language that wasn't theirs.
Large language models change the shape of that on-ramp. You can describe what you want in the language you actually think in, and the model can turn that intent into working code. A student in Kigali can describe an app for her community in her own words, watch it take shape, and then learn the concepts underneath by seeing what those words produced. The idea comes first. The syntax — and even the English — become things you grow into, not gates you have to clear before you're allowed in.
“Vibecoder” shouldn't be a slur.
Somewhere along the way the word turned into an insult — shorthand for someone who doesn't really understand what they're building. We think that's backwards. Directing an AI tool well — being precise about intent, noticing when the output is wrong, knowing what to ask next, and verifying the result — is a real skill. It's also the skill that lets a far larger group of people help build the future. Vibecoding is the front door, not the back door.
That is what this course lane is for: teaching computational thinking through plain-language direction first, then bridging to real code and real CS — so people the old on-ramp filtered out can walk in, build something true to their own vision, and go as far as they want.
