March 4, 2025, on the DT stage: Jeffrey Goldberg is skeptical. None of the arguments put forward by the two men sitting opposite him can convince him that AI could have a positive impact on art or society. The editor of The Atlantic Monthly magazine, as a fundamentally disagreeing moderator, provides a humorous contrast to his guests: authors Ayad Akhtar and Daniel Kehlmann, who have been friends for years and are engaged in a conversation about “large language models” such as ChatGTP, which they also test in practice. This three-way discussion, entitled Art, AI, and the stories we tell, was hosted by the American Academy, a venerable institution that welcomes American intellectuals from a wide range of disciplines as residents in a historic villa on Lake Wannsee.

Even though Akhtar and Kehlmann respond to the new technologies with curious and playful curiosity, they do not see the future of their profession as particularly bright. Akhtar already fears that he himself will become superfluous as an author in many fields, given that a large proportion of Hollywood and Netflix productions are already based on AI-based storytelling.

Overtaken by a technology 

As a European, Kehlmann hopes for regulation and predicts that human-made literature will continue to exist, but will become a luxury segment and only be printed in very small print runs. So the situation is serious for literature. Akhtar has tried out for himself how creative writing and AI might work together, and the result is his latest play, McNeal (translated by Daniel Kehlmann). To do this, he spent two years training a rapidly changing large language model until it was able to write like him: in his style, in his meaning. He says that he had to work extremely hard to extract individual sentences from his digital assistant. The play depicts an author named Jacob McNeal “in the near future” at the height of his fame (Nobel Prize for Literature) and at the end of his life (cirrhosis of the liver). And while McNeal is caught up in the destructive relationships and lifestyles of his past, he is simultaneously overtaken by a technology that enables him to write a high-quality first draft of a novel in just two days: “Seeing those pages come out of the printer was like watching the last iceberg sink off the coast of Greenland,” McNeal remarks before attempting to kill himself. And the longer the theater evening progresses, the more we suspect that AI has long since taken over the text and is preventing the main character from dying.

“The owl of Minerva begins its flight only with the fall of dusk,” remarks old Hegel in the preface to his philosophy of law. A sentence that Ayad Akhtar quotes on the podium at the DT. What Hegel means by this sentence has been interpreted in many ways: an understanding of social conditions is only possible after their reality has unfolded. The goddess Minerva is the Roman counterpart to the Greek Athena, and both goddesses, as guardians of wisdom, had the bird of wisdom as their mythological attribute, the nocturnal owl. Insight therefore only comes when something comes to an end; it concludes an era and does not establish it, as Hegel had previously claimed. It seems as if we have already entered a new era that will change our society more profoundly through AI than we are willing to admit.

When asked by Jeffrey Goldberg which art form is least susceptible to being taken over by AI, Akhtar and Kehlmann name comedians and theater actors. Akhtar has high hopes for theater as a place of resistance against the zeitgeist. He talks about the cave paintings in Lascaux, which were created more than 4,000 years ago and still speak to us today because they address a universal dimension of humanity. And people who stand on a stage and attract others with a story are actually a gift: the opportunity for living, fallible, inefficient, emotional, intelligent people to come together in a cave, as a theater can be.