Automation Bias

AI-infused applications are becoming incredibly good at “personalizing” our content, but will there come a time when we let algorithms make all of our decisions?

Automation bias refers to the tendency to favor the suggestions of automated systems.

Take Netflix, for example. Everything we see on the platform is the result of algorithms—even the preview images that are generated. Then, to harness the power of data and machine learning, Netflix categorizes its content into tens of thousands of micro-genres. Pairing these genre tags with a viewer’s history allows them to assign several of over 2,000 “taste profiles” to each user.

And while there’s nothing wrong with allowing Netflix to guide what we watch, there’s an enormous sea of content standing by. Estimates from 2015 claimed it would take nearly four years to watch all of Netflix’s content. Thousands more hours of content have since been added.

If we want to counter this cognitive bias, finding a new favorite series on platforms like Netflix may require some good old-fashioned human curiosity.