An artificial intelligence and its programmers teamed up and participated in a literary contest in Japan, almost winning the final prize.
It seems that 2016 will not only be the "year zero" for virtual reality, but for AI as well. Two teams entered submissions which were co-authored by human and AI writers. Earlier this month, DeepMind's AI computer defeated the world's best human Go player, four to one.
This shows that AI capabilities radically improved, and the AI writer further proves it.
One of the novels, aptly named "The day a computer writes a novel," ends with the following phrases:
"I writhed with joy, which I experienced for the first time, and kept writing with excitement. The day a computer wrote a novel. The computer, placing priority on the pursuit of its own joy, stopped working for humans."
Hitoshi Matsubara, a professor at Future University Hakodate, led the team that submitted one of the novels to the Nikkei Hoshi Shinichi Literary Contest.
The human side of the writing team chose the general guidelines for the novel, such as gender of characters and plot line. Then, the team selected a number of phrases, sentences and words that would fit the narrative arch. Finally, the AI was in charge of "writing" the novel by assembling the elements picked by the human team.
Satoshi Hase, a science fiction novelist, shared his positive impression during the press conference.
"I was surprised at the work because it was a well-structured novel," he notes.
However, he adds that the AI can improve on delivering character descriptions.
The Hoshi Shinichi Literary Award accepts writing that belongs to AI programs, but this is the first time when submissions coming from AIs enter the competition.
11 out of the submitted 1,450 novels were written by teams featuring humans and AIs.
The literary contest is structured into four parts, and each one filters the remaining submissions. The judges were not aware which novels had AI contribution and which were solely written by humans. Yet, one of the novels passed the first screening stage.
"So far, AI programs have often been used to solve problems that have answers, such as Go and Shogi," Matsubara said.
He goes on to add that he expects to see AI's potential being used in ways that replicate human creativity.
This poses an important question: should the AI keep developing, will it be able to mimic the "human mind" without understanding and incorporating its flaws? Check out this coverage which explains how Microsoft's Tay AI chatbot embeds the worst of human discourse.
