Everyone using AI has had the same experience: you type in a request that seems clear, hit Enter, and get back something that’s almost—but not quite—what you wanted. So you try again. And again. After a few rounds, you either get frustrated or you give up.

What’s really happening isn’t a technology failure—it’s a communication one. Generative AI is fluent, but not intuitive. It understands instructions, not intention. The people getting the most out of it aren’t necessarily the most technical; they’re the ones who’ve learned how to ask.

When generative AI first hit creative departments, it promised to change everything. Write faster. Design quicker. Concept in real time. For many teams, that’s partly come true—but the other part of the story is less shiny. What’s becoming clear is that AI doesn’t automatically make creative work better or easier. In many cases, it makes the gaps in process, clarity, and direction more visible.

The old programming adage “garbage in, garbage out” has never been more relevant. AI will dutifully generate exactly what you ask for—whether or not what you ask for is any good. That’s the uncomfortable truth now confronting both brands and agencies: the problem isn’t the technology. It’s the inputs.