Good morning from San Francisco.
I’m here at Workday Rising, the annual user conference of the enterprise cloud company that offers finance and HR solutions (Workday is a CFO Daily Sponsor). Yesterday, I had the opportunity to lead a discussion with Erik Brynjolfsson, director of the Stanford Digital Economy Lab and professor at the Stanford Institute for Human-Centered AI; and Ankur Agrawal, a partner at McKinsey’s corporate strategy and finance practice. Brynjolfsson and Agrawal shared insight on creating value with generative AI and the changing role of the CFO.
More access to data
Generative AI and large language models will give companies a way to use certain data sets they couldn't before, according to Brynjolfsson, who is also the co-founder and co-chairman of Workhelix, a startup that consults with companies on generative AI use cases. CFOs have always been at the forefront of working with data, he said. "And I suspect that they can help lead that revolution as we use these new kinds of data to make better decisions," Brynjolfsson said.
An exciting time to be a CFO
“I think the chief financial officer is now no longer a financial officer alone,” Agrawal said. “They’re asked to do a lot more. And that's part of the reason why the digital and technology evolution is so powerful because it allows them to enable the organization to do many things. CFOs are driving and shaping the agenda for companies. I would argue this is a very exciting time to be a CFO.” He predicts that more CEOs will come from the ranks of CFOs. Along with driving the strategy, data analysis is a strong skill for finance chiefs, Agrawal said. Boards that are looking for CFOs of the future are asking for those skills along with an aptitude for technology, like generative AI, he said.
Automation is freeing up time for functions like financial planning and analysis to really focus on creating value, Agrawal said. “Clients of mine who were not thinking about automation, are now thinking about it because automation is a prerequisite for enabling a lot more generative AI,” he said.
How to know whether generative AI will save money at your company
“We recommend that you don't just go by gut instinct, you can quantify this,” Brynjolfsson said. You can look at the occupations in your organization and break them each down to a set of tasks, he explained. You can give a score on whether there's a generative AI tool that can help with that particular task. For example, “helping with your email, yes, and loading boxes on the truck, no,” he said. And as you go through each of the different tasks, you basically come up with a score card and then assess what the organization needs, he said. “I wouldn't say it's the final answer,” Brynjolfsson said. “It's the beginning of a road map to tell you where there are really juicy opportunities, and where's the low-hanging fruit for our company compared to our competitors? Where’s most of the value going to be created?”
Getting employees on board with generative AI
“One of the methods that I found works very well is a hackathon,” Brynjolfsson said. “You tell everybody, ‘We’re going to take a day off from everything else you're doing, and we're going to show you some basic training, not just on ChatGPT, but a set of other tools that can help you. And we want you to use it for the particular things you're doing in your daily job.'” This unleashes employee creativity, and helps get the buy-in, he said. And in the process, there's a possibility an employee may even figure out how to use ChatGPT for an ad campaign or product design, for example, "in ways OpenAI hadn't even imagined," Brynjolfsson said. "Make sure they understand that the goal is not to replace the workforce with generative AI; it's to augment it," he said.
Sheryl Estrada
sheryl.estrada@fortune.com
