Over the past year, language models have entered the popular consciousness as new apps like ChatGPT attract hundreds of millions of users. Bold claims have been made about the capabilities and future impact of such products. Researchers in the field of natural language processing (the scholarly field in which language models originated) have been somewhat taken aback by these developments; language models have been around for decades. Noah Smith, an expert in natural language processing, takes to the Chicago Humanities stage for a presentation that aims to demystify generative artificial intelligence with a focus on language models. By introducing historical context and a few key concepts and foundational ideas, he aims to narrow the gap between scientific and popular understanding of language models while avoiding the hype.

• This program is presented as part of the NielsenIQ Series on Artificial Intelligence.
• Please visit our FAQ for more information about ticketing, books, accessibility options, and other event logistics.