- Do More With Less Using AI
- Posts
- A free deck on AI event graphics, and my session for event educators
A free deck on AI event graphics, and my session for event educators
Real outputs from Canva, code, and Higgsfield on the same fictional conference, plus my IMEX Frankfurt talk for the professors training tomorrow's event pros.
1. The AI graphics deck I didn't have time to teach
Last week we ran the intermediate session of Claude for Events with Julius Solaris. 130 event professionals and five demos in two hours. I had planned to touch on AI graphics inside one of those demos, and then the clock beat me. There was too much good content and not enough sessions.
So I built the segment out into a full deck instead.
I’ve been hearing more and more, "Which AI tool should I use for graphics?"
After a year of making event graphics with AI, I'm convinced that's the wrong question. A single conference creates more design than some brands make in a year, and no one tool covers it, because there are three fundamentally different ways AI makes graphics. Each wins at different tasks.

AI inside your design tool (Canva). The rawest-looking output of the three, but it functions as a real file your team can open and fix, and one approved template can autofill a finished card for every speaker on your roster.
AI writing the design as code. Every font, dimension, and badge number lands exactly where it was told to. 300 badges from a registration export is the same effort as 3, and a typo never gets printed 300 times. The trade is that every edit goes back through the code.
AI generating the imagery (I use Higgsfield through Claude; Midjourney and GPT Image are the same approach). It rendered a photoreal keynote ballroom that doesn't exist, in about 20 seconds. It still can't write a line of text inside an image without mangling it.
The deck shows real, unedited output from each approach for the same fictional conference, so you can see the differences yourself instead of taking my word for it. You'll also find pros and cons for each, the seven questions that separate them, and the workflow where all three stack on one project — a generated venue shot, made exact with code, handed to the team as a Canva template.
Grab it here: What's the Best Way to Make Event Graphics with AI → https://drive.google.com/file/d/1Sm8VM_HitKhbjcRefFfiuzXkro7ZCVA3/view
2. My session for the people who teach event pros
One more thing worth sharing. Back in May, IAEE and IMEX Frankfurt brought together the faculty who teach event management programs for a virtual session on where AI belongs in the curriculum. My segment demoed using ChatGPT to build assignments that train students to spot risks at events before they happen — the kind of judgment that usually takes a decade on show floors to develop, compressed into coursework.
The full recording just came out, with talks from Dr. David Jarman, Joanna Goodey, and Antoine Casanova-Mazet alongside mine. If you teach, train, or onboard event people, it's worth checking out.
Watch the session → https://drive.google.com/file/d/12OhHPU-zCmMs3mZE6g4VHTLHjIkuEFwI/view
Till next time,
Noah
Do More With Less Using AI
PS: If your team makes event graphics a different way, hit reply and tell me how.
Reply