Firm leadership in the age of AI

Charlene Li is a bestselling author, speaker, and executive advisor. With Katia Walsh, PhD, she cowrote the book Winning with AI: The 90-Day Blueprint for Success. On June 11 at AIA26, Li talked to an at-capacity crowd of architects, designers, and allied professionals about what it will take for their organizations to use AI in service of their goals. 

A gap only leadership can bridge

AI has unprecedented potential, but it can be a struggle to decide whether and how to chase that potential. As Li put it, “There’s a gap between knowing that AI is transformational and actually using AI to transform the way you work.” 

To illustrate the tension around AI, she brought up the example of FOMO (fear of missing out) to label people’s eagerness to adopt AI, then introduced the term FOGI (fear of getting in) to describe what’s holding organizations back. Li believes active leadership is necessary for bridging the “AI hesitancy gap,” letting organizations overcome their fear of getting in and avoid missing out.

However, per Li, leaders are sometimes reluctant to dive in with AI, telling themselves, “I’m the least qualified person to do this. I barely use technology, I barely know how to use AI.” But Li encouraged leaders to break that thought process because employees “are feeling a pressure to use AI … and there’s no answer for them. The firm doesn’t have a direction.” 

Shadow AI and pilot purgatory

In the absence of leadership setting a singular AI direction, employees will pick directions—plural—on their own. Different individuals and groups within an organization are likely to research and choose disparate approaches to AI “without it being fully sanctioned,” Li said,” leading to “silos of practices.” She referred to this as “shadow AI.” 

On a company-wide level, organizations also run into trouble when they try many AI adoption approaches without a coherent strategy. That leads to organizations “throwing things against the wall to see what sticks and doing tons of pilots, but they end up in pilot purgatory. This is the place where value goes to die,” said Li.

AI should support your business strategy, not become it

To avoid troubles like shadow AI and pilot purgatory, Li encouraged leaders to set an AI roadmap by focusing on their business (instead of AI) first. “Start with a place that you know so well: your clients, your business, your team … instead of just thinking about all the different ways that you can use AI.” She said organizations often “have a list of all the use cases” of AI, but “use cases are not the strategy.” 

Instead, Li said, “You have a strategy, so think about” how AI can help you carry that strategy out. To aid with that thought process, Li offered these questions for consideration:

• What do your best clients keep asking for that you can’t deliver fast enough?
• What does your team complain about every Friday afternoon?
• What would you love to offer if you had twice the capacity?

Building a roadmap

Later in the talk, Li contrasted short-term wins with the long-term planning that is necessary for AI adoption. “You have some quick wins—maybe some tools and designs that you can implement right now. But then there are some strategic things … that are going to take much longer.”

As a planning tool, Li recommended an 18-month roadmap. “To be clear,” she said, “this roadmap isn’t what you are going to do with AI. It’s a roadmap of value”—essentially, predicting where AI might support your business’s strategy and, as time passes, assessing whether those predictions were accurate.

Li acknowledged that 18 months is a long time, especially given how often technological advances can disrupt plans: “It sounds like madness to have an 18-month roadmap for AI when everything is changing.” But, she went on to say, “One of the things I’ve learned in studying the disruptive transformation is you have to plan for disruption, and an adaptive road map is [key for that].”

She continued by explaining the “adaptive” nature of the roadmap: “At a minimum, every quarter, sit down and open the roadmap, and you’ll understand” where your business started and how far it has come. After those quarterly check-ins, she recommended revising the next five quarters of the roadmap, in essence creating an 18-month rolling plan.

Adoption in practice

Li believes that firms should set policies that make employees feel comfortable using AI rather than constraining them. She used the example of Formula 1 racing, where drivers feel empowered to go fast because their cars have powerful brakes. “The whole goal here is to make sure people feel safe so they can overcome that AI hesitancy gap.”

She encouraged firms to adopt a blend of rules and flexibility. Quoting her coauthor Katia Walsh, she said, “Structure without flexibility is bureaucracy. Flexibility without structure is chaos.”

So, what can AI actually do for architects?

Near the end of her talk, Li zoomed in on the specific tasks AI might best help architects with. One example was creating lots of ideas at once; for instance, an architect might use AI to generate dozens of options for a site layout. “You can do that in minutes, and then decide these are the ones that make sense. You can bring clients into the process much earlier.” Since clients can’t always articulate exactly what they want, generating many models for them to look at can be helpful for picking a path they’re happy with. 

AI can also support architects by handling non-design tasks, like taking notes and creating summaries, that would otherwise draw their time and energy away from designing.

Getting started and moving forward

Regardless of the specific approach your firm takes, Li emphasized that it’s important to start the journey rather than wait until you have the perfect plan. “The minute you take that first step, you will move, and then you can make more decisions with more data.” By achieving some of your goals after you set out—even just 60% or 80% of them—you will learn “what the gap is between what you are doing and where you want to be.”

Danielle Steger is AIA's senior manager, editorial & publications.