Photo by Andrew Daddio
Despite the dire warnings that artificial intelligence will create major disruptions to the economy, higher education, and society as a whole, AI expert Avi Goldfarb remains optimistic about its potential.
“Clearly, AI is going to win, and how people are going to adjust is a fascinating problem,” Goldfarb told students and faculty members during his lecture “Power and Prediction: The Disruptive Economics of Artificial Intelligence,” sponsored by the Lampert Institute of Civic and Global Affairs — part of its spring 2023 series.
Goldfarb, Rotman Chair in artificial intelligence and healthcare at the Rotman School of Management, University of Toronto, acknowledges the potential changes coming due to AI are of particular importance to today’s college students, who will soon be joining the workforce.
Goldfarb said faculty teaching methods will have to adjust to accommodate the increased use of AI as more and more students turn to the technology to help answer questions.
“If we say it isn’t allowed, it’s favoring the students who cheat over those who don’t,” Goldfarb said. “We need to redesign how they do homework to change what they do.”
Goldfarb said students should think about what they’re trying to accomplish and how AI technology can be leveraged to help them achieve those goals. “I like to think that, because you’re here, your understanding of AI is deeper and your careers will be more successful because of this,” Goldfarb said, addressing the students in the crowd. “In 30 years, your success could be about your understanding of the disruptive economics of AI.”
Goldfarb said AI prediction is significant because it helps humans make better decisions, noting that solutions are necessarily incremental, and there will be a loss of a way we used to do things, including take-home assignments. “New tech will mess with the way we always did stuff, but it creates an incredible opportunity for you to interpret material in a way useful to you after you graduate.”
Goldfarb described what is currently happening with AI as the “between times,” in that it feels similar to how we viewed electricity in 1890, when we could see the potential of the technology but didn’t know exactly what the systems would look like. In addition to higher education, he predicted the medical, insurance, and transportation industries will all undergo significant changes due to the rapid increase in prediction technology.
While it may be terrifying to think about what jobs will be left for humans if machines can do everything we do now, Goldfarb said true artificial intelligence is still perhaps decades away. Right now, AI is based on prediction machines, allowing for better decision making, better work, and better organization, both faster and cheaper — such as calculating how likely someone will pay back a loan or making a medical diagnosis. Once the price of prediction falls, its possible uses will skyrocket, Goldfarb says, until it becomes a general purpose technology that transforms society, like the invention of electricity disrupted life for previous generations.
Goldfarb says that, as we deal with the anxiety of trying to understand what jobs are going to be replaced or changed, there will also be opportunities to specialize and upscale existing jobs, along with creating new ones.
“When a new technology comes along, it’s easy to think of all the things it will replace; if you make a living writing, especially if it’s pretty standard, you should worry about your job. On the flip side, tech is also going to create new jobs and opportunities, but they haven’t been invented yet, and they seem way off in the future.”
Max Edelstein ’23, a Lampert Scholar, said Goldfarb’s lecture was a fascinating discussion on how advances in AI will shape the future. “Professor Goldfarb’s description of artificial intelligence technology as ‘prediction technology’ was extremely helpful in clarifying the impact that this technology will hold,” Edelstein said, adding he plans to continue to experiment with ChatGPT and keep up with artificial intelligence technology as it evolves.