The problem with algorithms though is that they remove the outlier. The things that shape you are usually outside your normal scope of interest.
Professor of engineering at Oakland University Barbara Oakley was once a linguist until she realized she could apply the same “chunking” principles to become fluent in math. Mixing subjects broadened her understanding of how discovering new things work.
Algorithms never go deeper than the prescriptive answers. They take what’s most likely of interest and give you more of that, confirming your bias.
Human discovery is less fallible than machines. Aggregated tastes or wisdom of crowds is a viable recommendation engine. But the problem with people is a lack of time–we take too long to gather content and dig through it. The machines can sort through content streams faster, and with accuracy.
We can’t afford to our put our taste in any method. The only way to balance the curators, friend recommendation, with the algorithmic engines is to go manual, staying open to the possibility of discovering something outside our standards interests. Those magazines at the dentist's office are worth perusing.