It's easy to come up with new marketing use cases to implement with AI.
It's easy to experiment without a large budget or a lengthy technology project, so there is little friction in saying: let's just try it.
But, "easy to try" is a pretty low bar for deciding where a business should place resource.
When marketing technology projects required a significant effort, managers had to make a strong case for them before anyone even agreed to allocate an analyst.
With AI, you can be halfway into an experiment before anybody decides whether the use case was important enough to dedicate effort. And an experiment doesn't stay small just because it was cheap to start with!
Someone will spend time on it. That's the nature of organisational behavior.
Trying ten things because we can still means we need to choose where to spend time and money. The outcome should determine what we pursue, rather than what AI makes possible.