The Turing Test first passed in art?

In a 2019 article, I challenged the widespread belief among AI experts that creative professions would be the last to be replaced by AI. Later, in an interview, I suggested that the Turing Test—a measure of a machine’s ability to exhibit human-like intelligence—would first be passed in the arts.

The prevailing assumption has been that AI would master quantifiable data before tackling creative tasks. This idea clings to the outdated belief that the arts are the result of mysterious, uniquely human forces that reflect our consciousness and soul, setting us apart from other beings.

But reality tells a different story. Consider an AI facing three Turing Test challenges: A) Writing a scientific paper, B) Composing a journalistic op-ed, and C) Crafting a poem.

The poetry challenge would likely be the easiest to pass. Any AI-generated randomness could be interpreted by judges as creative choices, not errors. The op-ed would be tougher, since flawed opinions could be seen as missteps rather than creative expressions. But in scientific writing, any inaccuracies would result in immediate failure during peer review.

This example highlights not just where AI’s ability to mimic human intelligence may falter, but also how we, as humans, have failed to develop rigorous standards of excellence in the creative arts.