Ten years ago, highbrow culture magazine The New Yorker started a contest. Each week, a cartoon with no caption is published in the back of the magazine. Readers are encouraged to submit an apt and hilarious caption that captures the magazine’s infamous wit. Editors select the top three entries to vie for reader votes and the prestige of having captioned a New Yorker cartoon.
The magazine receives about 5,000 submissions each week, which are scrutinized by cartoon editor [Bob Mankoff] and a parade of assistants that burn out after a year or two. But soon, [Mankoff]’s assistants may have their own assistant thanks to Microsoft researcher [Dafna Shahaf].
[Dafna Shahaf] heard [Mankoff] give a speech about the New Yorker cartoon archive a year or so ago, and it got her thinking about the possibilities of the vast collection with regard to artificial intelligence. The intricate nuances of humor and wordplay have long presented a special challenge to creators. [Shahaf] wondered, could computers begin to learn what makes a caption funny, given a big enough canon?
[Shahaf] threw ninety years worth of wry, one-panel humor at the system. Given this knowledge base, she trained it to choose funny captions for cartoons based on the jokes of similar cartoons. But in order to help [Mankoff] and his assistants choose among the entries, the AI must be able to rank the comedic value of jokes. And since computer vision software is made to decipher photos and not drawings, [Shahaf] and her team faced another task: assigning keywords to each cartoon. The team described each one in terms of its contextual anchors and subsequently its situational anomalies. For example, in the image above, the context keywords could be car dealership, car, customer, and salesman. Anomalies might include claws, fangs, and zoomorphic automobile.
The result is about the best that could be hoped for, if one was being realistic. All of the cartoon editors’ chosen winners showed up among the AI’s top 55.8%, which means the AI could ultimately help [Mankoff and Co.] weed out just under half of the truly bad entries. While [Mankoff] sees the study’s results as a positive thing, he’ll continue to hire assistants for the foreseeable future.
Humor-enabled AI may still be in its infancy, but the implications of the advancement are already great. To give personal assistants like Siri and Cortana a funny bone is to make them that much more human. But is that necessarily a good thing?