Spotify’s supposed “smart feed” seems to be stuck on repeat, showing me the same over and over. One minute I’m looking for a niche artist, the next I’m drowning in a sea of the same ten songs, over and over again. Algorithms, these unseen forces that control what we see online, are shaping our visual experiences. And that’s not necessarily a good thing.
Two decades ago, discovering new tastes and ideas meant digging through book or music store shelves, leafing through magazines, and finding hidden gems on quirky blogs. Today, an automated system curates our online silence, but is it leading us to new things or just showing the same old stuff?
The problem lies in how hungry algorithms are for information. These digital beasts crave data, a constant stream of user preferences and clicks – a veritable buffet to fuel their recommendation engines. And if you’re a privacy-conscious individual like me, this presents a very big dilemma. To gain access to a few “recommended” posts, we have to relinquish a significant portion of our online preferences — and data.
I have a love-hate relationship with algorithms. I understand their (business) worth and potential to help us find new things, but the way they work with our data isn’t always great. They’re good at showing us what we already like, help us find similar stuff, but they don’t leave much room for surprises or happy accidents.
So you could say that it’s complicated.
As humans, we crave variety. Our interests are constantly changing, just like our personalities. And since algorithms feed on our past behavior, they have trouble keeping up with us, which doesn’t work well for curious minds who are always looking for something new.
What we really need is a smarter kind of algorithm, one that can connect the dots across the vast online world and different world views. What would a system look like that could predict our next artistic obsession based on all the things we click on and explore? That would be algorithmic curation with a twist!
But, there’s a catch. This dream algorithm would need a much deeper look into our online behavior (yikes), more information than most privacy-conscious people would be comfortable with. And all that data raises a whole new set of questions. Who gets to use it? How is it stored? Sold? Who owns it (the user or the company)?
We’re back to square one: are algorithms a gateway to discovery or a prison of predictable feeds? The answer, like most things, is probably a bit of both.
Afterword
Many of these thoughts have been percolating for a while now. But it was only in listening to a conversation between Ezra Klein and Kyle Chayka that I realized this is part of a wider problem that needs to be addressed in the long-term:
How can we solve this complex balance in order to fully benefit from this savvy technology?
Although discussions about paying users for their data keep vanishing in thin air (just one idea of many), I have the feeling this issue will increase in relevance as more and more people and governments critically question data practices and demand a more user-focused quid-pro-quo.