Wonder if NLP and machine learning algos are equipped to disambiguate the intended meaning of an autoantonym, with respect to its context.