Using this information, the model can then tell us the probability of a drug-protein interaction that we did not previously ...

Inaccurate metaphors about the mind can cause real harm, for instance leading educators to 'train' children like algorithms.