Have you ever wanted to ask a flight attendant, “what’s the weirdest thing you’ve ever seen a passenger do on a plane?” Okay, so maybe it’s not the most pressing inquiry you’ve ever wanted to make, but you have to admit, if flight attendants were asked this question, they would probably have some interesting stories to tell.
One of the great things about the internet is you can find the answer to this and just about any other question that has ever been asked. Quora is one website dedicated to questions and answers. People ask questions, others answer them, and other people can browse and read the answers. Like many websites, the Quora algorithms try to learn about their users so they can customize and personalize the content that appears in their feed.
I tend to read a lot about things related to the travel industry. So this question about weird passenger behavior popped up on my feed and it piqued my curiosity enough to click the link and read through some of the answers (riveting stuff, I assure you.)
Another article that piqued my curiosity was, “what would happen if the earth stopped rotating for one second and then resumed?” The great thing about Quora is you can find an obscure or highly theoretical topic and see how people who really know their stuff (astrophysics in this case) can explain it.
The problem (and bear with me while I explain why this is a problem) is that Quora has learned something about me. Quora now knows that I am curious about extreme astrophysical events and bizarre human behavior. So Quora suggests other questions I might be interested in, like, “what would happen if the earth got knocked off its orbit?” Or, “what’s the strangest thing you’ve ever seen somebody do in a restaurant?”
And these questions are interesting to me, so I click them too. And before you know it, Quora is showing me more and more content about highly improbable intergalactic events and weird human behavior. I shaped the algorithm, but now the algorithm is shaping me. Imagine over many years I become some kind of expert in astronomy and bizarre displays of public fetishism.
Either I let the algorithm shape me, or I have to push back and reshape the algorithm. I can’t base my reading preferences only on what I want to read in the moment; I have to think about what kind of reader (and maybe even what kind of person) I want to become.
So how do you train an algorithm? Here are a few ideas:
- Start from the real world. Search for things on the internet that you discovered from real world activities and conversations. This may help you break out of internet rabbit holes that can take you deeper and deeper down a dark spiral.
- Break patterns. When you notice the algorithm is feeding you a particular theme of content, try to break the pattern. If you keep seeing articles about weird human behavior, bizarre astronomy or political controversy, avoid those topics for a while and show the algorithm a different side of yourself.
- Be aspirational. Don’t just click on what interests you in the moment. Click on what you wish you were more interested in. Think, “what kind of person would I like to become?” And, “does my internet activity reflect my highest values?”
We are living in the age of the algorithms. They are always there, guiding us and nudging us, presumably based on our own expressed interests. But they get it wrong sometimes. And we might not realize that the world we are seeing around us is not the real world, but a highly filtered view of the world based on some version of ourselves that, let’s be honest, really exists. But it might not be the version that we want.