Will Artificial Intelligence ever figure out that imitation of something that looks credible is not the only way?