A few years ago, I read a book called The Dan Sullivan Question: Ask It and Transform Anyone’s Future. Its premise is that in world where everyone is competing to have the best answer, the more beneficial move is to ask a better question.

Sullivan’s question asks you to imagine that three years have passed from where you are right now. Whatever goal you're focused on has been completed and it’s been a resounding success. You're incredibly excited about your achievement. In fact, it couldn’t have gone any better.

With that in mind, you look back from that future and describe what must have happened, personally and professionally, for you to feel as happy as you do with your success. (You can read the actual question here.)

What Sullivan is asking you to do with his question is essentially time travel. I find this interesting because most planning begins in the present and moves forward, constrained by what we know today. Sullivan moves you into an imagined success and asks you to look backward, as though the future has already occurred. From there, you begin to see the decisions and changes that might have made it possible.

What he's doing isn't anything new and actually has a well established scientific foundation. Psychologists have studied related forms of prospective thinking and mental simulation for decades.

But what interests me about his question now, though, is how it can be applied to AI transformation.

Decide What Success Means

Before you can transform an organization with AI, it helps to understand what you are trying to transform it into. Many organizations already have an AI platform, a policy, and a growing list of pilots. They can describe what they are doing, but they struggle to describe the organization those activities are supposed to create.

Part of the difficulty is the speed of the technology. No one can describe with confidence what AI will be capable of three years from now—and if they are, they’re probably trying to sell you something. For AI transformation, I think eighteen months is a more useful horizon: it’s far enough away for work to change and close enough to demand you be specific in what you’re trying to become.

The Future in Three Views

Building on Dan Sullivan's question, I ask one of my own in AI Vision Workshops, where I have leaders imagine that eighteen months have passed and their organization’s work with AI has gone well. I then ask them to look at that future from three different points of view.

First, I start with the people doing the work. What is noticeably different for them? What has become easier, clearer, or less frustrating? Where do they have more space to exercise judgment and less need to search for information or reconstruct the context around a decision?

Then I ask them to consider the customer. What is noticeably different about their experience with the company? Are they receiving faster answers, better-informed service, or greater consistency? Has the business become easier to work with? Has it preserved the human relationship that customers value?

Finally, look at the future through the eyes of the leader. What can you see or understand more clearly? Which decisions can you make with greater confidence? How has your own role changed as the organization has changed around you?

The point is to make the future specific enough to examine. Leaders may struggle to invent a future from a blank page, but they often know when one feels recognizably like their company, when an ambition is too cautious, or when an attractive outcome would require the organization to become something it does not want to be.

Once that future is visible, it becomes a filter. It helps leaders decide which use cases matter, what a pilot must prove, what governance must protect, and whether a roadmap is building toward anything more meaningful than adoption.

Imagining the future forces leaders to decide their ideal state. AI expands what is possible. But leadership still has to choose what transformation looks like.