88718 : Bubbles (Big Data for Human Minds)

Fall 2017
Undergraduate Lecture, Social & Decision Sciences @ CMU
Scaife 219, Tuesdays & Thursdays 1:30 to 2:50 PM
Prof. Simon DeDeo / TA Zachary Wojtowicz (zdw[at]andrew.cmu.edu)
Zach’s Office Hours: Mondays, 11am–12pm in the SDS grad office (Porter Hall 321)
Simon’s Office Hours: after class, and otherwise by appointment (please meet with Zach at least once, first, before requesting an appointment)
Lecture Summaries

2017 Hackathon is live!

Pasted Graphic 1

Open discussions turn into echo chambers; optimistic traders pump good money into bad stocks; we fail to see, or sympathize, beyond the limits our culture, upbringing, or education prescribe. These bubbles—information bubbles, market bubbles, social bubbles—drive us to ask some of the most basic questions in the social sciences: Why do we believe the things we do? Where do our ideas come from, and how can we measure the consequences of their conjecture, spread, and evolution? How can we design systems to make us better thinkers? In this introduction to the “big data” study of human behavior, we'll learn some key concepts and simple computational tools for studying how people gain and share information, with a focus on what they say and write. And we'll apply these tools to social behaviors from the writing of Harry Potter fan fiction to online trolling, to science, markets, and liberal democracy itself. The class will include conceptual, computational, and data-driven investigations; students in social science, humanities, engineering and the sciences are equally welcome. At the end of this course, students will be able to build models for how people think and talk to each other, to see how thinking and talking work in both the past and present, and to imagine, and design, systems that might help us think, and talk, together more effectively in the future.

Pre-requisites: willingness and initiative to work with real-world data.


Week One: Bayesian Reasoning for Intelligent People / Test Your Knowledge (Fair vs. Biased Coins)
Week Two: Are Disagreements Honest? / Information Theory for Intelligent People
Week Three: Rules for Social Minds
Week Four: September Problem set
Week Five: Recipes for Science a.k.a., a cheat cheat to the 2017 Bubbles Hackathon / sprezzatura, PageRank, and the recursive cool