This study, led by the University of Pittsburgh, showed the participants poems written by ten famous English-language poets—Geoffrey Chaucer, William Shakespeare, Samuel Butler, Lord Byron, Walt Whitman, Emily Dickinson, TS Eliot, Allen Ginsberg, Sylvia Plath, and Dorothea Lasky—as well as poems generated by ChatGPT in a similar style.
The study's results showed that when it came to identifying whether a human or an AI wrote a poem, participants were more likely to misattribute an AI-generated poem as human-written than vice versa.
AI poems were judged to be human-authored roughly 75% of the time, compared to human-written poems, identified as human-authored significantly less frequently.
The participants ranked AI-generated poems higher in terms of overall quality than those written by humans. This finding goes against previous research suggesting a preference for human-authored works.
This study proposes that the preference for AI-generated poetry can be attributed to its accessibility and straightforwardness. Human-written poetry often involves layers of complexity and opacity, while AI poems communicate their themes, moods, and emotions more clearly and directly.
AI-generated poems may be more engaging for readers who need more background or interest in exploring the subtle nuances of more complex poetry because they provide instant clarity.
However, poet Joelle Taylor, author of the TS Eliot prize-winning C+nto & Othered Poems, expressed concern about the implications of this shift: ‘While AI is capable of generating technically proficient poetry, it lacks the core humanity that she believes defines true poetry.’
“A poem is more than an algorithm. It is meaning, empathy, revelation, inversion, dissidence, passion, and surprise”.
The study raises important questions about AI's ethical implications in the literary world. Because many readers have difficulty distinguishing between AI-generated and human-written poetry, it may be worthwhile for governments to consider implementing regulations for AI transparency.
As technology continues to advance, the line between human and machine creativity is becoming increasingly blurred, prompting a need for careful consideration of how AI-generated works are presented and understood.
As AI continues to evolve, it will be crucial to balance technological innovation with preserving the deeply human qualities that have traditionally defined the art of poetry.
