@vzn AFAIU it is randomized, adaptive algorithm, so "exponential speedup" is not true. Besides "In experiments, Singer and Balkanski demonstrated that their algorithm could sift through a data set which contained 1 million ratings from 6,000 users on 4,000 movies and recommend a personalized and diverse collection of movies for an individual user 20 times faster than the state-of-the-art."

20 times is indeed noticeable and superior (people fight over 1.23 vs 1.21), but far from mentionworthy clickbaity link it just was.

All it does is a kind of alpha-beta prunning for reading samples...

lol do feel the press release was a bit (over?)hyped, but we're talking harvard phds here... o_O bet if you look at the paper it will all make much more sense & seem far more reasonable... and anyway maybe no media picked up the PR anyway because hey its merely Computer Science :P

@vzn It should matter that this is PhD, but it doesn't. Have you ever seen wise person without diploma or not that wise with one? Never heard of academic points trickery via dubious journals? Yes, you have, you've told me. That is that...

You know, that it gets hard to take it serious, if at Harvards page in one paragraph there is “exponential” and in the next there is 20, nobody noticed? Nobody reviewed it? Authors do not care enough to look at page where theirs magnificent work is advertised?

If "ifs" and "buts" were candy and nuts, we'd all have a merry Christmas...

@Evil think you have valid points but apparently you still havent read the paper, am sure its more substantial than the blowing bubbles in the press release. complaining about hype in press releases is like asking why marketing/ advertising contains exaggerations.