[nosedrip]

some great books you're mentioning here, I'd love to see your recommendations in an app like this, perhaps we'd need to move away from an algorithmic recommendation system, but rather influencers based on 'moods'? There is a radio station that works like this NTS, they even claim "not curated by algorithms", but still scalable and solid music. I love this concept.

[lacker]

Yeah, the problem to me with algorithmic recommendations is that they aren't sufficient. I don't see that an algorithm recommended book X, and then immediately go buy the book. No algorithm is that good.

So if I see an interesting book from the Amazon algorithm, the natural thing is to read the reviews, but book reviews are really hard for me to trust. So many people reading the book have a different background, different priorities, different reading ability, they are just looking for something different in a book. So when I read book reviews I simultaneously have to figure out "does this reviewer think like me?" and "is this the sort of book I want to read?" And Amazon or Goodreads book reviews are just too shallow for this.

Over time, though, I start to learn what book recommenders I can trust. Many different sources. Some friends of mine, some bloggers, some book awards, they like the same sort of book I do and I respect their recommendations.

I read a lot of books, though, and I'm not really sure if the "mass reading public" is similar enough to me for my personal product judgment to be relevant to the task of building a Goodreads competitor....