What’s your favorite AI use case?
It’s the number one question I’m asked when I talk to people about AI. It’s also my least favorite. Not because it’s an inherently bad question, and not because my favorite use cases are hyper-specific to what I do and evolving on a near-daily basis. I don’t love it because it highlights an approach that has too many people stuck on the starting blocks.
Perhaps this sounds familiar. You have got a few friends or colleagues who won’t shut up about AI. You intuitively understand that the technology as a whole is a big deal, but you struggle to see how it maps to your specific situation and why so many people are obsessed with it. And every conversation where it comes up, far from leaving you inspired, only deepens the feeling that you’re missing something fundamental.
So, the hunt for obvious use cases ensues.
If you’re stuck in this position, the standard advice usually goes something like this: “Pick a specific problem in your business and commit to learning how AI can help you solve it.” In principle, it makes sense. I’ve given the advice to plenty of people myself. But I’m increasingly convinced that it’s wrong, or at least incomplete. And the reason so many people are stymied here is that they are approaching the situation from the wrong end.
Also by Jason Griffing: Forget Your Feelings
Contrast two different approaches to AI. Let’s call Approach 1 the practitioner’s method. The practitioner is all about applying AI, finding current challenges at work, and using the technology to solve them. It’s necessary and important work. But if your experience with AI is limited to the basics, this approach might be putting the cart before the horse. Without any intuition for what the technology is capable of in its more advanced (read: agentic) forms, it’s exceedingly difficult to see how it applies beyond the very basics.
Let’s call Approach 2 the exploratory method. The explorer is all about diving deep into the technology itself purely for the sake of learning. Solving a particular business problem through this process is great if the opportunity to do so is apparent, but that’s not the main goal. And, critically, the lack of a visible use case is not a reason to forgo the exploration. Exploration, by definition, isn’t about finding something you already know is there. It’s about possibility. The primary aim is simply to discover, to develop an intuition for precisely what tools are available and how they apply, and to viscerally feel both the capabilities and limitations of the technology.
In a recent conversation, Henry Clifford, an industry veteran and fellow AI nerd, made a comment that captures the importance of the exploratory method perfectly. He said, “This is an age that favors the curious.” The exploratory method can be hard for a busy leader or business owner to justify. Faced with a never-ending onslaught of imminent challenges in the business, there’s a strong temptation not to invest time in learning AI unless the payoff is clear. But that’s not what exploration is about, and waiting for an obvious use case with a clear payoff is exactly how you stay stuck.
Also by Jason Griffing: Differentiating in a Post-AI World
If you’ve bought into the concept that AI is a transformational technology, but unsure exactly what to do about it, you’re not alone, but you need to stop looking for the obvious answer. If you can’t think of a specific problem in your business that you want to go after, then do something else. Vibe code a game with your kids. Spend an afternoon setting up a tool like OpenAI Codex or Claude Cowork to do something simple like triage your inbox, manage your calendar, or schedule daily briefings. Try building an app to manage something simple, like meal planning or a to-do list.
The point isn’t the output. The point is the intuition you build in the process, and once you have it, the business applications become obvious on their own.
