ARTICLE | April 19, 2024 | VOICES

Embracing the contradictions of GenAI

To get ahead on a generative AI strategy, let’s first take a big step back

By Kevin Barnard, Workflow contributor

At the click of a button, applications powered by generative artificial intelligence (GenAI) can quickly make sense of massive amounts of unstructured data in the time it takes a human to get a cup of coffee. As a result, businesses are racing to adopt GenAI.

But building a resilient generative AI strategy is not as straightforward as it might appear.

By day, I speak to business leaders about resilience; at night, I practice and teach yoga. These may seem like disparate fields, but they have a great deal in common. For one, building resilience is a gradual process that necessitates daily discipline. It requires us to get comfortable being uncomfortable and to make peace with the counterintuitive. And lately, I’ve been thinking a lot about the discomfort that may arise as businesses charge full steam ahead on GenAI.

To avoid wasting resources and time, GenAI-curious leaders must embrace these three contradictions:

Generative AI

Executives are eager to create a generative AI strategy to apply to as many use cases as possible—as quickly as possible. And to be sure, many organizations can make their processes faster and cleaner by giving employees access to the technology as soon as they can.

But while rushing ahead, don’t neglect the fundamentals. Governance models and operating models need to evolve. In other words, don’t use yesterday’s models for tomorrow’s challenges. Without good AI governance, GenAI will produce outputs that are useless at best and harmful at worst. Counterintuitively, highly regulated industries such as finance are moving faster on GenAI than their less-regulated counterparts, according to research from firms such as McKinsey. These industries are mandated by law to keep track of the data they’re using—where it comes from, how noisy it is, and whether it’s producing outcomes that do harm. Because their data is cleaner, they aren’t forced to backtrack and make costly changes to algorithms that have been trained on bad data.

Gone are the days when a robot passing the Turing test made bold headlines. Now, our machines are making art and finding cures for diseases. The excitement is real and it’s warranted.

But slow down. The more we talk about what these powerful thinking machines can do for us, the more we need to talk about the humans who use them.

It’s tempting to start indiscriminately applying machines to as many problems as we can. However, start by thinking about the problems themselves and the people who must address them every day. Come up with a holistic, prioritized map of the challenges you’re trying to solve across the organization and carve those into small, winnable pieces. Micro-wins build momentum, develop cross-functional trust and, yes, drive a better employee experience.

To do that well, departments across the org need to get a lot better at talking to each other. What do they need? How can they get it? What sorts of conversations are necessary to facilitate these changes?

The onus is on executives to support their employees trying new things and possibly making a mess—and they shouldn’t have to worry that at the same time they’re learning how to leverage AI in their jobs they aren’t working themselves to the unemployment line. As tough as this might be, acknowledging the elephants in the room and collaborating on how to confront them will go a long way toward building a resilient generative AI strategy that stands the test of time.

With GenAI, suddenly every knowledge worker can do things faster and smarter, increasing their output and supercharging their creativity.

However, while it’s important to integrate new technologies into your company’s toolkit, it’s just as important to understand the tools you already own. Take stock of everything that’s on hand—for example, underutilized resources such as software licenses and apps. See how they all fit together. Then think about how you can layer GenAI on top of these resources. CFOs especially will appreciate a more cost-effective, pragmatic approach.

In these times of heady innovation, GenAI is a powerful enabling function. But it’s only that. It can’t tell people what to do—well, perhaps it can make recommendations! Ultimately, it is up to the humans in a company—managers, leaders, CXOs, directors—to decide how best to use these increasingly mesmerizing high-tech tools to guide their company to a more profitable future.

When I’ve led employees through massive changes, I’ve found that bringing humanity to the process emphasizes that we’re all in this together. Let’s not lose sight of our goal, which is to augment and enhance how humans operate in the workplace. We’re nothing without our people.