1 comment

[ 4.2 ms ] story [ 16.9 ms ] thread
They should really program "Darren" to produce something more like text written by a human being. Training a Markov chain on the output of your PR department, then having it write your "blog", isn't exactly effective.