Benedict Evans in his article “The death of the newsfeed” very clearly provides insight on why big tech companies are forced to a more algorithmic world, inadvertently forcing them between a rock and a hard place.
I couldn’t agree more to all Benedict states in this article and highly recommend anyone interested in this subject to read the article (twice!).
Unavoidable as it seems, though, this approach has two problems. First, getting that sample ‘right’ is very hard, and beset by all sorts of conceptual challenges. But second, even if it’s a successful sample, it’s still a sample.
Using signals of what people seem to want to see risks over-fitting, circularity and filter bubbles. People’s desires change, and they get bored of things.
In my opinion this is one of the biggest challenges in data science.
How can we help people navigate an abundance of choices (post, pictures, search results or products) without unethically steer behaviour, creating self-fulfilling prophecy1 or limit the room for people to change their preferences and discover new stuff.