With the advent of online business and the rapid advancements in artificial intelligence and automated computer systems has come an entirely new method of interacting with customers. Whereas before, prices, products, and experience was largely uniform across customer segments, new technology allows internet companies to create a customized user experience by learning from and reacting to a customers purchase, viewing, and even search history.

A prime example of the power, for good or ill, of the learning algorithm is YouTube. Everyone who watches many videos on the famous video content streaming site has had the experience of a very interesting video popping up in the suggestions that they would otherwise never have found. However, YouTube also illustrates a serious flaw of automation and provides a warning to those seeking a similar mechanism: YouTube’s algorithms are infamous for their inaccuracy and lack of fairness when it comes to things like copyrighted material or inappropriate content, often demonetizing perfectly innocent videos while leaving others untouched.

Another instance of personalized interactions is Amazon. Items left in an Amazon user’s cart for long periods of time will often have their price slightly reduced in hopes of stimulating a purchase. Almost every online sales business will also give recommendations of products that a particular user has expressed interest in… Or that match similar profiles’ interests. At times it can be uncanny how a site suggests something you would normally have not thought of, but still interests you.

Internet businesses benefit greatly from the ability to connect with and personalize the experience of their customers, but they must be very careful to be sure they do not over-automate their systems, or risk their buyers becoming leery of the inhuman nature of their website and advertisement.