>Add the wrong ingredients into the Big Data recipe, and what you’ll get is a potentially toxic output.

Shouldn't the Big Data algorithms apply the same principles?