In Defense of Recommendation

Amen!

There is much more content out there than any one person can experience, people need something that can filter it. Before we had algorithms we had "tastemakers". Back in the day of Casey Kasem, the "Top 40" was not derived from sales figures but rather what was being played at elite clubs and by elite radio stations.

(Once real sales figures got out, they shocked the industry. Country music turned out to be a lot more popular than people thought in 1985, but the Country music industry itself struggles with not accepting its own best sellers being crossover artists such as Garth Brooks and Taylor Swift)

I have been experimenting with an RSS reader that does content-based recommendations after thinking... We've had the technology to do this post 2000, why is it we are still stuck in this bad place with collaborative filtering recommendations?

I can say I understand why a bit better, I think that collaborative filtering works better as most people would define "better", particularly when the goal is to filter the top 1% as opposed to filtering the top 20% of a feed. Still, content-based filtering is a way to put yourself in control of algorithms... And w/o control of algorithms you are not in control.