Not something you look at. Something that works. An interactive tool, a calculator, a quiz, a simple website. Something you can click on, type into, and share with other people. And you're going to do it without writing a single line of code.
This is possible because of a category of AI tools that has exploded in the last couple of years, sometimes called "vibe coding" or no-code AI app building. The idea is simple: you describe what you want in plain language, and the AI writes the code to make it. You see the result immediately. If it's not quite right, you describe what you'd like to change, and the AI updates it. Back and forth, refining as you go, until you have something useful.
The term "vibe coding" was coined by Andrej Karpathy, a well-known AI researcher, in early 2025. He described it as an approach where you "fully give in to the vibes, embrace exponentials, and forget that the code even exists." For non-technical people, this is even more straightforward: you don't need to forget the code exists, because you never had to think about it in the first place. You just describe what you want, and it appears.
This matters for professionals because it means that small, useful tools that would previously have required a developer (or at best, a complex spreadsheet) can now be created by anyone. Need a calculator that works out project costs with VAT? A checklist that tracks progress across multiple categories? A simple quiz to test knowledge after a training session? A unit converter for a specific niche? These are all things you can build in a conversation with AI, often in under thirty minutes.
The results aren't going to compete with professionally developed software, and they don't need to. They're what some people call "disposable apps": small, purpose-built tools that solve a specific problem. Think of them less like buying software and more like building a really good spreadsheet, except the output is something interactive that looks and feels like a proper application.
What "vibe coding" is and why it matters
Throughout this programme, you've been a consumer of AI outputs: reading what it writes, looking at what it generates, listening to what it creates. In this Thing, you become a builder. You move from asking AI to produce content to directing AI to create functional tools. That's a real shift in the relationship between you and the technology.
For professionals, the immediate practical value is in creating small, purpose-built tools. Every workplace has processes that live in spreadsheets, checklists that exist as Word documents, and calculations that people do manually or in their heads. Many of these could be turned into simple, interactive tools. Not to replace enterprise software, but to fill the gaps between the big systems.
The limitations are real, though. Vibe-coded tools are good for simple, self-contained applications, but they're not a replacement for professional software development. They can break in unexpected ways, they're difficult to maintain over time, and they're not suitable for anything involving sensitive data, critical business processes, or complex logic. Think of them as the AI equivalent of a hand-drawn poster rather than professional graphic design: perfectly good for many purposes, but not for everything.
The skill you'll practise here, clearly describing what you want an AI to build, evaluating the result, and iteratively refining it, is one of the most transferable AI skills you can develop. Whether the tools are called vibe coding platforms, AI app builders, or something else entirely in two years' time, the ability to translate a clear idea into a working tool through conversation with AI is likely to remain useful.
What tools are available
There are several tools that let non-technical people build interactive applications from natural language descriptions. Things are moving fast in this space, but as of early 2026, the most accessible options are:
All of these tools work in your web browser. None of them require you to install anything or have any technical knowledge. The core skill is the same one you've been developing throughout this programme: describing what you want clearly enough for AI to act on it.
Resources to explore
Written by one of the tool providers and naturally favours their own product, but it's a useful overview of the tools available and what each is designed for.
A clear explanation of the concept from Google, including how their tools fit into the picture. Good for understanding where this category sits in the broader AI space.
The primary tool for this Thing's activity. Sign up for a free account and ask Claude to build you something interactive. Artifacts appear in the right-hand panel of any conversation.
A dedicated AI app builder with a free tier (five daily credits for public projects). Good for building more complex applications that need multiple pages or a real URL.
Another AI app builder with a limited free tier. Browser-based, no installation needed. Results can be deployed as working websites.
Google's AI app builder, currently free during its preview period. No credit limits on the free tier, though the tool is newer and less polished than the alternatives.
Searching YouTube for "build an app with Claude Artifacts" or "Lovable tutorial 2026" will turn up video demonstrations. Watching someone else go through the process of describing an app and iterating on it is helpful before you try it yourself; you'll pick up techniques for how to describe what you want effectively.
Activity: build a useful tool with AI
This activity asks you to build a simple interactive tool using nothing but natural language descriptions. You'll use Claude Artifacts as your primary tool (since it requires no sign-up beyond a free Claude account), and optionally try a second tool for comparison.
-
Choose what to build. Pick something small, specific, and useful to you. The best projects for this activity are tools you'd actually use, even simple ones.
-
Build it in Claude Artifacts. Open a new conversation at claude.ai and describe what you want to build. Be specific. The more detail you provide about what the tool should do, how it should look, and how someone would use it, the better the result will be.
-
Iterate and improve. Almost certainly, the first version won't be exactly what you had in mind. This is where the iterative process begins, and it's where the real skill lies.
-
Try a dedicated builder (optional). If you'd like to see how a dedicated app builder compares, try recreating the same tool (or building something new) in Lovable, Bolt, or Firebase Studio.
-
Reflect on the experience. Step back and think about what just happened. You described something in plain language, and a working interactive tool appeared.
Your output
Your output for this Thing should include:
- A screenshot or link to your completed tool (if using Claude Artifacts, you can share the conversation link; if using Lovable or Bolt, you can share the deployed URL)
- The initial prompt you used to describe what you wanted
- At least three examples of refinement prompts you used to improve the tool
- If you tried a second builder, a brief comparison of the two experiences
- A short reflection covering what worked, what surprised you, and what you'd build next
Why this matters
The ability to build working software from natural language descriptions would have seemed like science fiction five years ago. Today, it's something you can do on a free tier in your lunch break.
For professionals, the immediate value is in creating small, purpose-built tools. Every workplace has processes that live in spreadsheets, checklists that exist as Word documents, and calculations that people do manually. Many of these could be turned into simple, interactive tools, not to replace enterprise software, but to fill the gaps between the big systems.
The skill you've practised here (clearly describing what you want an AI to build, evaluating the result, and iteratively refining it) is arguably one of the most transferable AI skills you can develop. Whether the specific tools change or not, the ability to translate a clear idea into a working tool through conversation with AI is likely to remain valuable.
Claim your Open Badge
Submit a screenshot or link to your completed tool, your initial prompt and at least three refinement prompts, your optional comparison of two building tools, and your reflection on the experience.
Submit your tool, prompts, and reflection to claim this badge via cred.scot.
Claim now