Künstliche Intelligenz: Arbeit neu denken

The limits of poetry-writing machines

Slate just published an article on Eureka, a machine that attempted—and somewhat exceeded—to automate the production of poetry. Rather than being a curio and a sideshow of the time, Eureka was instead, “The Victorians loved a show, especially one that drew upon both science and illusion for an unforgettable performance. The Eureka offered both. The machine was a feat of industrial progress, but sensational enough to be entertaining. And it can still wow. The unveiling of a recently restored version of Clark’s invention drew crowds of excited visitors who, like spectators past, reveled in the show.”

But can a machine ever really produce poetry? Well, an interesting take on it was presented by Oscar Schwartz, at a TED Talks event in Sydney, Australia. Schwartz, who co-created the Bot or Not website, said that any poetry produced through machines reflected the input—that if you put Emily Dickinson into a machine and asked it to produce a poem, it would produce an Emily Dickinson-like poem. Quite simply, you get out what you have put in. This, says Schwartz, raises interesting questions about how we define what it is to be human.

Ultimately, poetry-generating machines fail. Not because they fail to produce poems but because they fail to produce anything outside the narrow lines it is given. We can teach a machine to do something, but they’re not yet at the point where the machine can learn it by themselves.

But perhaps the most-telling line of the Slate piece is its first one: “Poetry allows us to explore the depths of what [it] means to be human.” A poem from a machine cannot resonate like Leonard Cohen writing to a dying love; connect us like Raymond Carver’s ‘Late Fragment’; draw in political and social context as Muhammad Ali—probably inadvertently—with, “Me./We.”; or raise a wry giggle like Roger McGough.