Are we making our own decisions or are machine learning algorithms making them for us?

Kartik Hosanagar, author of the book, A Human’s Guide to Machine Intelligence: How Algorithms are Shaping Our Lives and How We Can Stay in Control, explains that algorithms are merely a set of steps for making decisions. Yet he points out that artificial intelligence has become so pervasive in our lives that we’re often unaware of just how many decisions machines are making on our behalf: “The algorithms [are] driving 70-80% of the choices that people [make]…[But] if we asked people how much of your choices are driven by algorithms, they might say maybe 10-20%. We think we are…choosing…but in reality, they are curating our world for us.”

In this interview, we talk about what companies should be asked to reveal about their algorithms. We also discuss why we need to educate ourselves about how they work. We also discuss some of the unexpected research findings that arise when machines learn from each other, rather than humans. For example, in one study, a surprising thing happened as machines were learning how to negotiate: Karthik explains that “…the bots were negotiating with each other using words and sentences that made almost no sense to the researchers. The bots had figured out a secret code to communicate with each other that was allowing them to communicate more efficiently.”

Kartik Hosanagar is Professor of Technology and Digital Business and Professor of Marketing at the University of Pennsylvania’s Wharton School of Business. His writing has appeared in Wired, Forbes, and the Harvard Business Review.

Episode Links

@KHosanagar

Irresistible by Adam Alter

Reporter Carole Cadwalladr’s The Guardian article on Google search bias regarding Jews

A more recent article on Google search algorithm bias

Kevin Gibbs and the Google autocomplete origin story

Code of Hammurabi

Jennifer Logg and her work on algorithmic and human judgment

Berkeley Dietvost, Joe Simmons and Cade Massey’s paper on our how humans avoid algorithms after they make errors

Rene Kizilcec’s paper on the effects of transparency on trust when it comes to algorithms

James Barrat

You can learn more about Curious Minds Host and Creator, Gayle Allen @CuriousGayle and www.gayleallen.net.

If you enjoy the podcast, there are three simple things you can do to support our work. First, subscribe. That way you’ll never miss an episode. Second, tell a friend or family member. You'll always have someone to talk to about the interview. Third, rate and review the podcast wherever you subscribe. You’ll be helping others find their next podcast.

A Short List of Places Where You Can Find Curious Minds

Spotify

iTunes

Tunein

Stitcher

Google Play

Overcast