Press

By the time companies are discussing bailouts, "they're already toast, and you're just trying to resurrect them from the dead," said professor Deborah Lucas. Bailouts can have a bad reputation, she said, but there are some cases where there is some justification. That's why, Lucas added, there was a lot of support for bailing out the banks in response to the 2008 financial crisis.

Senior lecturer Robert Pozen wrote: "Private credit has opened new borrowing channels to companies that otherwise struggle to access capital markets. But now that private-credit and secondary funds have sold shares to millions of retail investors, the pricing of loan transactions should be more transparent."

AI's "potential for job destruction is very significant, and those who hold capital will get richer," said professor Simon Johnson. For him, what's needed is "worker-friendly AI," a technology that enhances human capabilities, accelerates the acquisition of expertise, and creates new tasks, instead of rendering human skills obsolete.

Professor Thomas Malone and co-authors have developed a new way of analyzing work, to help predict which jobs are most vulnerable to AI disruption. "Instead of focusing primarily on jobs," said Malone, "we focus on the detailed activities that people and machines do at work." AI mainly threatens workers who manage information. But not all of them. Malone noted that some industries demand human empathy, a sense of ethics, and a knack for teamwork.

"The rhetoric out there is that the tools are going to be democratizing. But the reality is that you require a certain degree of education, abstract and quantitative skills, familiarity with computers and coding in order to be using the models," said Institute Professor Daron Acemoglu. "AI is going to increase inequality between labour and capital."

Research, by assistant professor Mert Demirer, associate professor John Horton, Peyman Shahidi (PhD candidate), and co-authors models production as a sequence of interdependent steps and shows that it is this interdependence that determines the true extent of the gains enabled by AI. "We are seeking to understand the effect of AI at the overall system level, not as a one-off productivity tool applied on a task-by-task basis," said Shahidi.