For many people, ChatGPT appeared to come out of nowhere — a startling wake-up call to the potential opportunities and threats of artificial intelligence (AI). But Zsolt Katona, professor of marketing at California’s Berkeley Haas School of Business, was using a forerunner, GPT-2, as early as 2019, both to teach executives on the technology leadership programme about emerging tech and to write scripts for the videos that accompanied his course.

“It was a very tedious task to write scripts that I’d then read from a teleprompter,” recalls Katona. “I’ve never been good at writing nice-sounding text, either in English or my native Hungarian. But, even back then, the scripts generated were pretty good.”

Katona has little time for ChatGPT naysayers. “I love it. It’s a fantastic educational tool,” he says. “I remember what it was like in high school when there was no internet. Just like Google Search for a previous generation, ChatGPT will become how people access knowledge in this generation. It means we can be much more efficient in education, including executive education.”