(Good) poetry can hone AI-HI skill sets

Anne Carson's Nobel Prize in Literature should be celebrated for reasons extending well beyond literature. At a time when AI threatens to make human intelligence (HI) seem an increasingly quaint biological accessory, poetry offers a remarkably effective way of sharpening the latter. The future belongs not merely to those who can operate AI, but to those whose HI, hooked to AI, can extract something exceptional from the partnership. Reading good poetry may prove to be an unexpectedly useful sharpening tool.

Unlike prose, which usually insists that meaning travels in the straightest of lines, poetry makes language work overtime. Words acquire multiple identities, phrases resist immediate interpretation, and nuance becomes almost all. Readers must dismantle language into its atomic constituents, examining semiotic possibilities of individual words alongside the psychological effects of rhythm and cadence. This is cognitive cross-training: pattern recognition, interpretative flexibility, sustained attention, and sensitivity to ambiguity.

Cognitive research on literary reading and figurative language suggests that such engagement can recruit processes associated with inference, mental simulation and reinterpretation. These distinctions matter in an age when AI can manufacture passable verse in seconds. Much of it is slop, as is an overwhelming amount of human-produced poetry: sentimental statements chopped into lines mistaken for profundity. The point isn't to worship verse but to cultivate discrimination, and spot and understand nuance. Humans who train themselves to detect the finest possibilities in poetry may remain harder to replace armed with this valuable skill set. Oh, and poems are usually short - like Roger McGough's '40-Love' in the pic above - perfect for our impatient times.