Photography Meets AI: Designing an Erasmus+ Course for Visual Educators

The Room

Twelve educators from eight countries. Some had never opened an AI tool. Some had never seriously studied composition. All of them needed to leave with skills they could bring back to their own classrooms.

This was the brief for the Erasmus+ course I designed and led — integrating AI into photography and visual content creation. Practical, not theoretical. International, which meant every assumption about shared context was wrong.

I spent more time designing this course than any photography project I’ve done. The photography was the easy part.

The Two Traps

Most AI photography courses fall into one of two traps.

The tool trap: here’s how to use Midjourney, here’s how to write a prompt, here’s how to upscale. The problem is shelf life. Tools change every six months. A course built around specific tools expires in weeks.

The theory trap: here’s the ethics of AI art, here’s the authorship debate, here’s where the technology is heading. Participants leave with opinions but no skills.

I designed for neither. Instead, I built the course around a single question: when does AI make visual work better, and when does it make it worse?

That question doesn’t expire.

What I Actually Built

The structure started with photography, not AI.

Day one was composition, lighting, intentional framing. No screens. No software. Just cameras and available light. This felt counterintuitive for an AI course, and some participants said so. But you can’t evaluate an AI-generated image if you don’t know what makes a good image in the first place.

The foundation had to be visual literacy. Everything else built on top.

Days two and three introduced AI — generation on one side, editing on the other. Where AI generates well: concept visualization, mood boards, content where the idea matters more than the proof. Where it falls apart: documentation, anything that needs to be real, anything someone will scrutinize closely.

Day four was integration. How to build workflows that use AI where it helps and traditional methods where they’re needed. Not either-or. Both-and, but with judgment about which tool fits where.

Day five was about teaching — how to structure these concepts for different age groups, how to handle the ethics questions that inevitably arise.

The Thing I Didn’t Expect

The most important insight came from watching the participants work, not from the curriculum.

Educators who understood composition and light used AI tools to extend their existing abilities. They recognized when a generated image looked subtly wrong — the light fell at an impossible angle, the depth of field didn’t match the focal length, the shadows contradicted the highlights. They couldn’t always articulate why it was wrong. But they could see it.

Educators who jumped straight to AI without the photographic foundation produced more content but worse content. Speed without judgment.

This pattern repeated every day. Visual literacy came first. AI fluency was useful only on top of it.

I’ve noticed the same thing in my own work. The camera teaches you to see. AI teaches you to produce. Seeing must come first, or you’re producing with no way to evaluate what you’ve made.

What They Took Home

The most useful thing wasn’t technical — it was a decision framework.

Is this task about documentation or creation? Documentation needs a camera. Creation can benefit from AI. Does authenticity matter? If yes, photograph it. If the concept matters more than the proof, consider AI. Can I evaluate the output? If you can’t tell whether the AI’s result is good, you shouldn’t be using it yet.

Simple questions. But they require the visual literacy to answer honestly.

What I Took Home

The course confirmed something I’d suspected: the future of visual education isn’t photography versus AI. It’s developing the judgment to use both well.

The educators who got the most from the week weren’t the most technically skilled. They were the ones who already understood visual storytelling and recognized AI as a new instrument in an existing practice.

Not a replacement for the practice. An addition to it.