With the smell of the crisp yellow corners of a sticky note, everyone in the room is acutely aware of the opportunity to stretch our brains, to flex our creativity, to collaborate in an inherently human way. Not only was the invitation to the sprint an opportunity to set our status messages to do not disturb for the week, it was an invitation to think without expectations. No matter how wild the ideas got along the way, at the end of the week the structure promised a deliverable worth envying.
Earlier this year something happened. Our deliverable was, well... bad. It wasn't even roasted — that would require a point of view someone didn't agree with. It was deflated, soft, mush. The audience was underwhelmed. I wrote it off as an anomaly. Have I lost my edge? Did I miss a step? I moved on. And then it happened again.
AI is a solution. It's a piece of technology that can be used to address a human problem. Design sprints are structured to find solutions. In the field of AI, we are now given the solution and asked to find the problem.
Design sprints are structured to find solutions. In the field of AI, we are now given the solution and asked to find the problem.
AI products, although shiny and bursting with potential, are untethered to the needs of people. We see it in the apprehension of new users, in the product discussions that churn around an undefined problem, in the data that shows we haven't yet found product-market fit. We've built something incredible and can't yet figure out how to apply it.
So what does this mean for our sprint? Is this still a method we can rely on? It needs to be. There is no more pressing time for deep, human, design thinking than now. The models will move forward with an inertia that feels out of our control. If we can solve this problem — and solve it at scale as a repeatable process — then maybe we can offer a lighthouse of application to steer solutions toward good.
Perhaps a sprint to solve the problem of a broken sprint. Will report back on progress.