12-05-2025
How effective project management is revolutionizing AI transformation in organizations
For many businesses, AI is a must-have. Whether integrating it into customer service functions or using it to cut down busy work and boost productivity, companies now see AI as an essential tool in their technology arsenal. But how can leaders ensure that their companies are using it to its fullest potential?
That was the topic of a recent panel discussion sponsored by the Project Management Institute (PMI) at the Fast Company Grill during SXSW. Three business leaders weighed in with their thoughts on how companies should—and shouldn't—proceed when jumping into the world of AI. Here are three takeaways from their conversation. (Scroll to the bottom to watch the entire panel discussion.)
1. Think big. Start small. Iterate often.
In the startup tech world, the mantra has long been 'move fast and break things.' But when it comes to AI, it pays to take a more calculated and deliberate approach. Kathleen Walch, director, AI engagement and community, Project Management Institute, sees a much higher failure rate when companies roll out AI projects compared to typical tech upgrades. 'What is causing these projects to fail?' she asked. 'It really comes down to making sure you understand what problem you're trying to solve . . . and making sure that AI is the right solution.'
That doesn't mean companies should play it safe when considering where to apply powerful AI technology. 'The other challenges we see with organizations is not dreaming big enough,' said Sarah Bird, chief product officer of responsible AI at Microsoft. 'You really need to pick that right use case where, when you get it right, you're going to be absolutely thrilled with the outcome. If you start with something boring and then it turns out boring, you're like, 'Why did we go through all that effort?' '
Leaders should also beware of the potential for a bureaucratic watering-down of their ambitious AI plans. That's been the experience of Tom Geraghty, vice president of digital innovation at Universal Destinations & Experiences. He's been on the team building Universal's Epic Universe, the highly anticipated new theme park that's set to open in May at Universal Orlando Resort. 'Some of my best friends are lawyers, finance people, and they're all going to chip away at your idea,' Geraghty said. 'So, make sure that your vision is super grand and you have room to compromise for physics and gravity, lawyers and money and time, and all the rest of that stuff, and you'll end up with something amazing.'
2. Make your team smarter.
Using AI effectively is a learned skill. Before rolling out access to AI throughout the organization, companies need to take the time to educate their employees up and down the ranks on its potential—and pitfalls. 'The most successful organizations that we've seen start with that first and figure out how they're empowering and educating their employees,' Bird said. 'And then you start to see the magic happen after that.'
That on-the-job training doesn't have to be handled in-house. Companies such as PMI provide e-learning resources to help upskill and educate employees across an organization. At PMI, professionals can access a broad menu of education and certification programs, with many aimed at helping project managers learn how to use generative AI and gain insights into prompt engineering. 'Everybody's at a different stage in their AI journey, and that's absolutely okay,' Walch said. '[PMI] is being very forward-leaning and saying, 'How can we help those project professionals along their entire journey?''
3. Get out of the sandbox.
The first step in introducing AI is identifying where and how to apply it. Once that decision is made, the next phase is to get that project off the ground quickly. Walch noted the strong temptation to stay in the sandbox—to run tests over and over until there's certainty that it'll work in the real world. 'We always say do pilots, not proof of concepts,' she said. 'Proof of concepts prove nothing.' To ensure AI solutions perform effectively and scale successfully, AI project leaders should prioritize real-world pilots with real users, enabling performance monitoring, feedback collection, and immediate issue resolution for optimal production readiness. Geraghty noted that NBCUniversal is constantly running pilots of AI projects it is working on, whether for theme parks, television, or movies. He and other leaders at the company play the role of test subjects, but those projects also make their way to consumers as a way to help NBCUniversal gain insights into how they're working. Ultimately, however, the goal of those projects isn't to make users notice AI. Instead, it's to use the power of AI to create something that truly wows them. At Universal Studios, that means using AI technology to build a new kind of immersive theme park experience. 'What we're trying to do is make guests forget about it,' Geraghty said. 'It should just be about the magic and the fun.'
Watch the full panel:
