It seems like a silly question: Can a computer write poetry? But then, after writing a poem, will that computer then take over the world? Oscar Schwartz is less concerned with the speculative danger surrounding artificial intelligence – instead, he’s using algorithms to help us figure out what it means to be human. In this intriguing talk from TEDxYouth@Sydney from May 2015, Schwartz (who by the way is hilarious on Twitter) discusses his Bot or Not project. The project was started as a website he and his colleague Benjamin Laird created that tests users by having them guess whether or not a poem was written by a human or a computer algorithm. Running these tests, he’s encountered some unique results, and some insight into what people collectively define as “human.” It’s a talk filled with fascinating ideas that are both technical and philosophical, so check it out for some brain-bending and thought-provoking speculation about the future of artificial intelligence.
Usually we post books for our recommended reading, but so far, Schwartz hasn’t written a book – just a handful of really interesting and frustratingly short essays mostly about people on the internet. The essay above seems somewhat inspired by the same questions he asks in his TED talk, but is concerned with a mysterious historical “computer” called “The Turk” and its modern-day counterparts.
The Most Human Human: What Artificial Intelligence Teaches Us About Being Alive, Brian Christian (Anchor Books, 2012)
In this book, poet Brian Christian tells the story of his participation in the 2009 Turing Test, where judges had to engage in five-minute instant-message conversations with unidentified partners, and then guess whether the person on the other end was actually a person, or a computer. It’s another interesting take on the Turing Test, artificial intelligence, and what being “human” actually means.