The Many Futures of ChatGPT

See the latest issue of Briefings at newsstands or read in our new format here.

By: Peter Lauria

No one on the marketing team made the connection. But somewhere out in the academic world was a reference to an old slogan that aligned perfectly with a new product campaign. Before the team even knew it, it had a go-to-market campaign that presented the product in a new and exciting light. And to think: if not for ChatGPT, it never would have happened.

In the half year that ChatGPT has been available, nearly every business leader has become familiar with this generative artificial-intelligence platform. After all, some 13 million people are already using ChatGPT every day, delighting in its uncanny ability to create original content indistinguishable from that produced by humans—and to carry out, in minutes, research that previously would have taken hours. But it is the corporate world that has the most at stake, as it seeks to capitalize on this groundbreaking; technology. “No one expected it to capture the business world’s attention so quickly,” says Tamilla Triantoro, a Quinnipiac University business analytics and information systems professor who studies artificial intelligence.

Yet few firms have found the perfect use for the technology, and given how this and other versions of AI work—by “learning” as it receives more information—that isn’t surprising. Jamen Graves, global coleader for the CEO & Enterprise Leadership Development practice at Korn Ferry, says the application’s potential is “mind-boggling,” but that the technology is evolving so fast that future uses are hard to predict.

On the most basic level, companies are leveraging ChatGPT to automate simple tasks like writing and updating code, responding to basic customer-service requests, summarizing lengthy texts like legal precedents, and creating templates for job descriptions or press releases, among others. But experts say the uses of the technology can go far beyond that. Who needs an entire team of engineers at a firm when ChatGPT can write customer and website code? Or an inside group of financial analysts when the tool can game out the pros and cons of stock buyback or ways to cut costs.

“Instead of just producing things, it can interpret them,” says Mark Esposito, cofounder of AI startup Nexus FrontierTech and an instructor at Harvard University. With ChatGPT, operations teams can better assess risks, while sales can better find patterns in customer feedback and recommend next steps. Esposito envisions more and more advances—for instance, he suggests ChatGPT could serve as a tool to aggregate different data points on ESG measures and generate a scorecard for investors.

5 Milestones in AI History

As artificial intelligence enters into its next great evolution, here are five seminal moments that got us this far.

1950:

Claude Shannon builds a robotic mouse named Theseus that can remember its way through a maze.

1987:

AI system NetTalk learns to pronounce some English words by matching them to phonetic descriptions. 

2012:

An AI system called AlexNet identifies pictures of cats, dogs, and various objects at near human levels.

2021:

AI continues to enhance the ability to generate images from text, as well as provide recommendations based on searches, social-media posts, and past online purchases.

2022:

ChatGPT and other AI systems become capable of writing original creative text comparable to that produced by humans.

To be sure, experts say firms do need to proceed with caution. Some major financial-services firms, for instance, have banned its use to protect against inadvertent violations of regulatory rules or client privacy. Other companies face similar risks, since whatever is transmitted to the new AI stays with it. More broadly, ChatGPT and systems like it flat-out make mistakes and aren’t immune to the same racial, ethnic, and gender biases that have plagued previous AI platforms.

Still, the growth potential is hard to deny. Daron Acemoglu, an economics professor at MIT, says smart companies are looking at how new versions of AI can be best teamed with workers to complement human output. Put another way, the more humans use ChatGPT, for example, the more analytical, creative, and value-added it will get. That’s part of the reason firms are now hiring “prompt engineers” and others who can optimize interactions with ChatGPT and produce the most relevant content. The better firms get at using such systems efficiently, the more opportunities they will find.

Which—in the end—leaves the many futures of AI in the hands of the C-suite and the boards they answer to. How fast companies move, say experts like Esposito, depends on whether leaders view the technology as a strategic function or just an additive to strategic functions. “It’s the difference between envisioning AI as running the company and supporting how it runs,” he says.