Modern day oracles or bullshit machines: how to thrive in a ChatGPT world

Talk by Carl Bergstrom. This is a talk inspired by this course: thebullshitmachines.com

Abstract

Large language models (LLMs) have upended education. Students and faculty alike are struggling with the pace of change. We have developed a general education course, Modern Day Oracles or Bullshit Machines, for every college freshman and high school student wanting to reflect on what it means to be human in an LLM-infused world. The course takes a collaborative learning approach to a fundamental question grounded in the humanities: how can we learn and thrive with LLMs?

In this talk, I provide an overview of the course, discuss what we have learned about teaching undergraduates to think about what it means to be a learner, scholar, and human being in a ChatGPT world, and reflect on the challenges of teaching about this rapidly evolving technology.

Course website: thebullshitmachines.com

About Carl Bergstrom

Carl Bergstrom is a Professor of Biology at the University of Washington. Broadly, his research contributions are many, ranging from how evolution encodes information in genomes to how it moves and distorts across online networks. Bergstrom’s many research contributions include the Eigenfactor metrics for measuring the influence of scholarly journals, foundational work on the science of science and the incentives that shape research, and models of how misinformation spreads through social media and, more recently, through generative AI. He is the co-author of the book Calling Bullshit: The Art of Scepticism in a Data-Driven World. Bergstrom is an eager science communicator, and many will recognise Bergstrom from his communication effort during the COVID-19 pandemic, when he gained a wide following. Trained in evolutionary biology and mathematical population genetics, he earned his B.A. from Harvard and his PhD from Stanford, and his work continues to cross the boundaries between the natural and social sciences.