By Shandana Mufti

We take technology for granted. We expect low prices on goods purchased through Amazon, and we expect cars requested through Uber to show up promptly. We think less about the algorithms that make this technology possible and about the impact of the thousands of algorithms that make decisions, big and small, about our lives every day.

“We’re surrounded by these algorithmic systems,” says Christo Wilson, a CCIS professor. “Some of them are innocuous – Netflix recommends stuff to us and we say, “OK great.” Others are hugely consequential, like are you going to get this job, are you going to go to jail?”

Wilson’s current work digs into algorithms we interact with every day, and he aims to build tools that can dig into and understand any given algorithm system. He also plans to look at the convergence of online and offline tracking, and how loyalty cards offer discounts but also mine user data, showing customers targeted ads online based on their in-store purchases. This research is supported by a CAREER award granted by the National Science Foundation for a total of about $500,000 dispersed over a period of five years. While the overarching goal is to create a set of general tools and methodologies that could be applied to any system, Wilson named several companies in his proposal, including Uber and Amazon.

With Amazon, Wilson wants to dig into the challenges faced by online marketplaces. On Amazon, many items have a button that allows customers to buy that item with a single click. Different sellers can set different prices for the same item though, and there is no guarantee that you will pay the lowest price – the seller you buy from is determined by an algorithm.

“You’ve got sellers, you’ve got customers, in the middle you have a bunch of data,” Wilson explains. “Do you use that [data] to help customers? Do you use that to help sellers? Do you use it to enrich yourself? Who knows what’s going on, and what is good and what is bad. It sort of depends on your perspective.”

Wilson has investigated Uber before, digging into the mechanisms of surge pricing. Now, he’s considering exploring how ride sharing services interact with existing transportation infrastructure like taxis and public transportation. “Maybe ride-sharing is good. We all share the same pool of vehicles,” he says. “Maybe they’re taking ridership away from things like busses or trains that are much more efficient but not quite as convenient.”

Proposals submitted for a CAREER award must include an educational component, and Wilson’s is a far-reaching one. He wants to address the lack of ethics training for computer scientists, and make it an integral part of the curriculum. While he’s considered creating a specific course around data ethics, Wilson hopes that ultimately, an ethics component might be incorporated into all computer science courses.

Such a course would ask students to think about the data they collect, why they collect it, and whether people’s privacy is respected in the process. “Is what you’re doing ethical?” Wilson asks. “People don’t get that kind of training, and I really think that they should.”