June 29, 2015
Scientists recently built a deep learning machine that outperforms the average human on IQ tests. But could an artificially intelligent computer write “Shall I compare thee to a summer’s day” or “The Three Questions” or become the world’s highest paid DJ? In other words, could a robot write “I, Robot”?
Computers haven’t yet replaced Shakespeare, Tolstoy or Tiesto, but Dartmouth College today announced a global contest for AI algorithms that create “human-quality” fiction, poetry and DJ dance music sets. The goal is to see whether human judges – in the form of literary readers and party-goers on a dance floor – can determine which creations are generated by machines and which by humans.
Dartmouth’s Neukom Institute for Computational Science will host the first annual Neukom Institute Prizes in Computational Arts throughout the 2015-2016 academic year. The “Turing Tests in Creativity” will include three different competitions -- DigiLit, PoetiX, Algorhythms -- for algorithms that create “human-quality” short stories, sonnets and dance music sets.
The competitions aim to inspire innovations in computational methods that generate artistic products, such as literary, musical and visual art. The Turing test, named for the British pioneering computer scientist and mathematician Alan Turing, tests a machine’s ability to exhibit intelligent behavior that is indistinguishable from that of a human.
“The classic Turing test has always been held out as the benchmark of human intelligence -- that is, writing an algorithm that enables a computer to carry on a conversation that fools the human on the other side,” says Dan Rockmore, a professor of Mathematics and Computer Science and the director of the Neukom Institute. “But even with a test like that, people always say a computer can’t write a novel or a poem. In our competitions, we hope to inspire artificial intelligence researchers to take on that challenge and create another dimension of AI -- creative intelligence.”
The Neukom Institute is collaborating with Dartmouth’s Digital Music and Sonic Arts programs on the DJ competition, which will include algorithm finalists and expert human DJs performing at a dance party at the Dartmouth Digital Arts Exposition.
“With services such as Spotify, Pandora, Apple Genius, Google Music Mix and others, we’ve become accustomed to algorithms mediating our enjoyment of music and other media,” says Michael Casey, a professor of Music and Computer Science. “But can we tell the difference between humans and machines when it comes to selecting the music we want to hear? Can a machine manipulate party-goers on a dance floor like a human DJ? The goal of the competition is to find out whether it’s possible for algorithms to be preferred to humans in musical performance.”
More information is available about contest registration, rules, prize money and deadlines at Turing Tests in Creativity at https://math.dartmouth.edu/~turingtests/