Generative AI and Finance

Artificial Intelligence has exploded in popularity in the past decade, so what can we expect to see with the development of ‘generative’ AI in the coming decade in finance? This talk explores the impacts of AI on firms, products, and labor, as well as potential uses for generating data. If we define AI as a “prediction machine,” meaning it consists of algorithms that make predictions from data and can make decisions, that also means it can subsequently be used to ‘generate’ and produce new content. ‘Generative’ AI can be useful in processing large datasets and performing more varied tasks. Will we see a democratization of access because of programs like ChatGPT or maybe we will see increased concentration as some firms disproportionately benefit from, for example, owning more computing resources? Does the use of AI benefit firms? Does it have the same effect on large and small firms? Does it lead to more automation of human tasks? Why has AI led to growth and product innovation, but hasn’t had any effect on productivity? Also, what are the potential applications for researchers? To what extent can generative AI be used to replace surveys, to pilot experiments/surveys, and to potentially uncover underlying mechanisms for why people rate some things positively whereas have negative expectations about other things? In general, is the use of AI a good or bad thing, and what effects will it have on people and the economy? These topics are explored by panelists Sendhil Mullainathan, John Horton, and Erik Brynjolfsson. Friday, Jan 03, 2025 2:30 pm - 4:30 pm PST Yerba Buena Salon 9 CHAIR: Anastassia Fedyk (University of California, Berkeley) PANELISTS: Erik Brynjolfsson (Stanford University), John Horton (Massachusetts Institute of Technology), Sendhil Mullainathan (University of Chicago) 0:00 Anastassia Fedyk, University of California, Berkeley 11:31 Sendhil Mullainathan, University of Chicago 36:44 John Horton, Massachusetts Institute of Technology 49:31 Erik Brynjolfsson, Stanford University 1:09:14 Q&A