How are companies really using AI?

Generative synthetic intelligence is transitioning from a curious toy to an indispensable instrument that extra corporations are integrating into each day operations, in keeping with a brand new report by Wharton Faculty specialists and advertising and marketing consultancy GBK Collective.

The report “Growing Up: Navigating Gen AI’s Early Years” is a survey of greater than 800 senior enterprise leaders in massive organizations that reveals a seismic shift of their attitudes and functions of AI in simply a short while.

In 2023, the primary yr of the survey, solely 37% reported utilizing AI weekly. That quantity has risen to 72% in 2024. Unfavorable perceptions, specifically fear and skepticism, are softening as decision-makers discover how this evolving expertise will help their companies change into higher.

“I was quite pleased to see the results. The picture we are getting is that people are learning about gen AI—and what they are learning, they like,” mentioned Wharton advertising and marketing professor Stefano Puntoni, who can be school co-director of AI at Wharton. He wrote the report together with Jeremy Korst, companion at GBK Collective, and Mary Purk, govt director of AI at Wharton.

Based on the survey, generative AI is being extensively deployed throughout capabilities, even in departments corresponding to advertising and marketing and human sources that have been initially slower to undertake it. The very best use is in doc and proposal writing and modifying with 64%, adopted carefully by knowledge analyses and analytics at 62%. Different high-use capabilities embody customer service and help (58%), fraud detection and prevention (55%), and monetary forecasting and planning (53%).

Puntoni mentioned the survey additionally exhibits the hole is closing between the 2 excessive beliefs about AI being both a harbinger of doom or a gateway to utopia. As folks change into extra acquainted with it, they’re much less curious and fewer fearful. They’re extra keen to experiment with AI whereas understanding its limitations, together with false information often known as hallucinations. Survey contributors cited accuracy, bias, knowledge privateness, workforce integration, and ethics as their prime considerations.

“You should think of it as a tool, and tools are not intrinsically bad,” Puntoni mentioned. “The danger is that you come to accept anything it says, and we need to build in systems to fact check.”

Wanting forward

Puntoni mentioned the survey is usually constructive about the way forward for AI, however he didn’t wish to speculate about what the identical questions would possibly reveal 5 years from now. Nonetheless, he is positive about one factor: Organizations will preserve experimenting with AI to determine the way it works finest for them. And because it evolves, so will the opposite expertise required to make it extra useable and sensible, corresponding to voice interface.

“I don’t believe the hallucination problem is ever going to be completely solved, but it’s clear we are making massive progress,” he mentioned. “There’s a lot of gas in the tank in terms of learning how to use this technology and developing complementary innovation.”

Though AI has been in improvement for many years, Puntoni and his colleagues agree that the speedy developments of the previous few years have ushered in a brand new period, one through which the transitions will come a lot quicker than, say, the 40 years it took factories to maneuver from utilizing steam engines to utilizing electrical energy.

“This is only the beginning,” Puntoni mentioned. “I think the big rewards are not going to come from doing things better; the big rewards are going to come from doing better things.”