From Data To Profit

CHAPTER 10The AI Maturity Model

Over the next two years, companies will make the same mistake over and over again. As AI hype continues to grow, companies will start looking for shortcuts to go straight to AI, like a race car going from 0 to 100.

With a race car, you'll hear it go through gears, revving the engine and then powering down. On the outside, all we see is a car flying forward. If we were trying to reverse engineer the motor, we might never understand the power of gears just by watching it drive.

It's the same thing with going from digital to AI. There are few shortcuts, and even those require a framework. Every business, product, and initiative goes through a maturity model. In past iterations, startups have had the advantage of not being bogged down by an existing solution. Even startups building from the ground up still have to go through a progression to get from digital to AI.

It's a common mistake. Startup founders want to build machine learning solutions because they are cutting-edge. They'll call in a data scientist to evaluate the potential. Most of us have been through something like this. We've all devised creative ways of explaining that without data, there's no way to build a machine learning solution.

First, the startup needs access to something that generates data. Without a version 1 digital product, the startup doesn't have a way forward to gather data and build a model. Without that original digital product, the startup doesn't even have someplace ...