How classic digital transformation lessons apply to AI—and what’s different this time around

Speaking to an audience of a few hundred business leaders during a webinar hosted by MIT on LinkedIn recently, he shared the most critical skills for leaders ushering in AI transformation; what organizations are underestimating; and what they need to do differently to succeed with AI.

His main takeaway, however, is that success with digital transformation has nothing to do with the technology itself. It all hinges on how an organization approaches the change. 

AI is moving too fast for you to plan

“The idea of planning and executing a five-year-plan, it just doesn’t work as well in this time when technologies are changing rapidly, when customer expectations are changing rapidly, and when the competitive market is changing so fast,” Westerman said. 

Given the breakneck pace of AI development, he believes companies instead need to develop a “much more emergent process.” The idea is that it doesn’t matter how long you spend planning, because you can’t predict what the landscape will look like in the far or even near future. Instead, set a vision for where to go, and then continuously help the organization move its way in that direction.

For leaders, this means first getting comfortable with the idea of “directive emergence,” and then remaining grounded in outcomes, not tools, according to Westerman. He suggests developing feedback mechanisms to assess how transformation is going.

“Carve up smaller projects so you can learn fast, maybe fail fast, but move quickly,” he said. “And always make sure that every step you take, you have an idea of what you’re expecting to get, so then you can figure out whether you got it or not. That’s really critical.”

From vision to motivation

Setting the vision is paramount because it’s the precursor to motivating people toward it, which Westerman said is the most critical nontechnical skill for leaders ushering in AI transformation.

For those at the top of the organization, he said, they’re concerned about the costs associated with AI: risks, accuracy, and compliance. At the bottom, people are just wondering if they’ll lose their job and how they’ll adapt. He described it as decision-making inertia at the top of the organization versus adoption inertia farther down. The challenge is moving both along and bridging the gap. 

“Helping to create the case for change and helping people feel that they can be part of that change. That’s becoming even more critical this time around,” he said.

Measuring the impact of this, however, isn’t so straightforward. He suggested looking at communication. Are the messages resonating? Are people changing in the culture to get there?

Finding wins will help move people along, and it’s tempting to start with the low-hanging fruit. But Westerman has seen companies underestimating the low-hanging fruit, and it’s important to keep expectations in check. 

“It’s not really as low as we think it is, because what we do on the bench in the lab is a well-behaved environment, and the world is not well behaved,” he said, referring to what one executive told him of their experience.

So, where should organizations steer their focus when it comes to operationalizing AI: optimizing tasks or deploying new functions?

“Choose your favorite framework. But what I would say is, if the framework is do everything at once, that’s not the kind of framework you want,” he said. “What you want is a framework that’s going to have some of the technology side, some of the organizational side, and a built-in process to do learning while you are at it.”