Three years ago, producing visualisations like this for a global congress campaign would have involved a long relay race between agencies, specialists and production teams before anyone could properly evaluate whether the creative direction even worked. Even in late 2024 during Eurovision planning these capabilities weren't really ready.

A strategist might define the positioning, a creative director would sketch initial thinking, designers would build moodboards, an experiential agency would begin spatial concepts and a 3D team would eventually render the booth weeks later after multiple rounds of interpretation and compromise.

Now that gap between thought and visualisation has narrowed dramatically. Not perfectly, and certainly not without friction, but enough that a single creative lead can explore campaign worlds conversationally and visually in near real time, testing atmosphere, emotional tone, architecture, media placement and experiential ideas while the strategic thinking itself is still evolving rather than after it’s already calcified into decks and timelines.

That distinction matters because most conversations around GenAI still focus on speed, automation or replacement, which honestly feels like the least interesting part of the shift.

The more meaningful change is that visualisation no longer arrives at the end of the process. It now participates in the thinking itself. This fictional campaign started with a made-up pharmaceutical product called ALOTTEXX™, positioned around travel-related motion instability because travel sickness occupies an unusually useful emotional territory. Universal enough to recognise immediately, uncomfortable enough to feel real, but lighter than the heavier emotional framing most pharma campaigns understandably sit within.

The strategic territory emerged quickly: movement versus calm. The world moving around you while stability quietly returns internally.

The line followed naturally: “The world moves. You can too.”

Traditionally, that line would then disappear into briefing documents while multiple teams attempted to visualise it independently. Instead, the image-making process started immediately alongside the thinking itself.

The first visual centred on a father and daughter asleep together inside a train or aircraft cabin while movement continued around them outside. No symptom diagrams, no exaggerated emotional performance, no glossy “healthy lifestyle” clichés, just stillness amidst motion.

What became interesting very quickly was not the generation itself, but the iterative judgement process around it because the AI constantly pushed toward excess: over-lit streets, impossible transport geometry, hyper-cinematic reflections and the strange polished perfection that still gives many generated visuals away instantly.

The creative work was often in removing things rather than adding them.

Less branding.
Less explanation.
Less spectacle.
More atmosphere.
More restraint.

Eventually the image stopped feeling generated and started feeling observed, which is a very different threshold entirely.

The tram image pushed the exercise further because it transformed the campaign from a simple media placement into something more environmental and cinematic. Once the congress branding disappeared and the setting became a believable European high street rather than an exaggerated activation zone, the image suddenly became more sophisticated. The campaign no longer felt inserted into the city. It felt embedded within it.

That distinction became important throughout the process because GenAI has a constant tendency to over-explain itself visually. Every render wants to become a presentation slide. Real campaign photography rarely behaves like that. The strongest images leave space for the viewer to complete part of the narrative themselves, which is partly why the balcony perspective worked so well in the final version. Instead of feeling like a media mock-up, the scene started feeling more like a captured moment, people noticing the tram passing below almost accidentally because it briefly transformed an ordinary city street into something cinematic.

Again, the interesting part wasn’t image generation itself. It was creative direction. Recognising when realism became more powerful than spectacle, when partial visibility became more elegant than total visibility and when a city should feel uneven, cluttered and lived-in rather than perfectly composed.

Taste became the differentiator far more than prompting syntax.

The experiential render revealed an even bigger shift because historically ambitious congress environments have been constrained by visualisation bottlenecks almost as much as budget. Exploring spatial ideas properly takes enormous time when every iteration requires interpretation across strategy, architecture, experience design and production.

GenAI changes that dynamic because architectural thinking, audience flow, emotional positioning and campaign systems can now all be explored together while the concept itself remains fluid enough to improve.

The final booth direction centred around a suspended kinetic ribbon sculpture flowing above an immersive “Experience Stability” platform where visitors physically experience subtle instability while cinematic transport motion gradually resolves into calm around them.

Importantly, the strongest direction only emerged through repeated simplification because the AI constantly tried to turn the booth into a futuristic theme park: too many screens, too many gimmicks, too much symmetry, too much corporate language, too much “future”.

The better version became calmer. Less exhibition stand, more architectural experience. The renders themselves weren’t really the outcome. They became thinking tools; ways to test emotional coherence, audience behaviour and experiential logic before a production team had even entered the room.

That’s the real shift underneath all of this.

Not AI replacing creative people.
Not magic prompting tricks.
Not infinite automated content.

The real shift is that the distance between strategic thought and visual articulation has shortened dramatically. You can now explore visual systems while ideas are still evolving, identify weak emotional logic before production hardens around it and prototype campaign ecosystems before budgets consume flexibility.

Ironically, this probably makes human creative judgement more important rather than less because once generation becomes abundant, discernment becomes the scarce skill.

And perhaps that’s the most interesting thing GenAI has quietly changed: creative people are becoming dramatically more visually articulate while ideas are still alive enough to evolve.