I think the original quotation simply used it as a rhetorical device to indicate difference in approach at a more abstract level than literally replacing conditionals with Bayes filters.