Depending on where you look, taste-anticipating algorithms have been either the 'answer' or the blunt instrument for solving the problem of information superabundance.