Sponsored Content
A parent’s account of Cresconova Labs’ Microorganisms course
I’ll be honest: when another parent mentioned an education centre in Chelsea teaching children to work with AI, my first instinct was to roll my eyes. We hear enough about AI doing children’s work for them. The last thing I wanted was to pay for my daughter to sit in front of a chatbot and call it learning.
Then I visited Cresconova Labs. Then I watched what actually happened.
The Microorganisms course doesn’t begin with AI at all. It begins exactly where good science should: at the source. Students prepare their own microscope slides, examine real microbial samples, and sit with what they see long enough to form genuine questions. What are these organisms? How do they move? What would they need to survive? These questions weren’t provided by the lesson. They were drawn out by it. The children earned their curiosity before they were handed any tools.
Only then, once real knowledge existed, did AI enter the process.
The task was this: take what you now understand about your chosen microorganism and prompt the AI to help you design a macro-environment in which it could thrive.
That framing changes everything. The students weren’t asking AI to do their thinking. They were testing their own thinking against it. And to do that well, you have to understand something first. My daughter had to know enough about her organism to recognise when the AI’s suggestions made sense, and when they didn’t. She had to evaluate, push back, and redirect.
This is the heart of what Cresconova Labs calls Human + AI Interaction. The human always comes first. The framework moves students through five deliberate steps: Question, Frame, Unpack, Own, and Iterate.
Question teaches them the difference between a surface prompt and a meaningful one, that a lazy question gets a thin answer, while a layered, precise question draws on the full depth of what AI has absorbed from human knowledge.
Frame trains them to build the prompt with care: what context to include, what to constrain, where their own assumptions might quietly shape the output.
Unpack is where it gets genuinely interesting. AI produces fluent, confident-sounding answers. This step teaches children that fluency is not the same as accuracy; that every output carries the fingerprints of its training data, including the gaps and biases. Students learn to ask not just what the AI gave them, but what it left out. Whose perspective is missing? What would this look like trained on different data?
Own is a commitment: I evaluated this, I decided this, I stand behind this. The AI cannot be held responsible. Humans can always.
Iterate closes the loop, refine the question, improve the frame, and go again.
What this produces is children who understand when to reach for AI and how to interrogate what comes back. Not passive users. Critical thinkers.
After the AI phase, students built their environments physically, by hand, using the Labs’ fabrication tools. My daughter’s model sat on her desk for weeks. She’d made it. Not in the thin sense of having typed some prompts, but genuinely: the ideas were hers, the judgment was hers, the decisions were hers. AI had been a collaborator, the way a reference book is, except requiring considerably more critical thinking to use well.
That’s the point. In a few years, every room my daughter walks into professionally will involve AI somewhere. The question isn’t whether she’ll use it. It’s whether she’ll know how, whether she can spot a plausible wrong answer, ask a better question, and own what she decides.
Cresconova Labs is building that capacity right now. I came in sceptical. I left wishing the course had existed when I was twelve.
