Large Language Model forms of artificial intelligence, like ChatGPT, don’t behave like the machines of years past, business professor Ethan Mollick writes. Interacting with them, you feel that you’re dealing with something profoundly new — a “co-intelligence” — that will radically alter industry, education, creativity, and more. General purpose technologies like generative AI are akin to the invention of the steam engine or the internet, Mollick explains. And while no one knows exactly how generative AI will develop or change humanity, this book offers an intriguing overview of the possibilities.
Today’s AI operates like an “alien mind.”
People have been fascinated by the idea of thinking machines for centuries. The first mechanical chess-playing machine was introduced in 1770, though it turned out to be a scam. In the 1950s, information theorist Claude Shannon and mathematician Alan Turing articulated the theoretical foundations for artificial intelligence. MIT computer scientist John McCarthy coined the phrase “artificial intelligence” in 1956. Researchers successfully programmed computers to solve logic problems and play games like checkers. However, progress quickly stagnated. Over the years, AI has gone through many similar cycles: rapid leaps forward, usually accompanied by technological advances, followed by periods in which little happens — “AI winters.”
“Supervised learning” uses annotated data, labeled with the right answers for achieving a desired output, to teach an AI system to, for example, recognize human faces or to predict which products you’re most likely to buy. This approach requires vast amounts of data — labeled images of faces or customers’ browsing histories.
If you want...
Ethan Mollick is a professor of management at the Wharton School at the University of Pennsylvania. He specializes in entrepreneurship and innovation. He has written for Forbes, The New York Times, and The Wall Street Journal.
Comment on this summary