Major Advancement In Artificial Intelligence : AI Algorithm Can Now Identify Humor
There is no doubt that Artificial Intelligence has left no stones unturned in the past years and has emerged extraordinarily well. With the continuous progressions in AI and its grounds laid in almost every area, Artificial Intelligence has now made apparatuses that can identify humor.
A remarkable differentiating factor between us humans and robots/machines is humor, an attribute which is purely humane. However, scientists and researchers have now come up with an algorithm which is knowledgeable enough of both identifying and fashioning humor, a major advancement towards creating “common sense” machinery.
Big thanks to Arjun Chandrasekaran, a scientist from Virginia Tech who has made this algorithm that can apparently identify humor now.
Mr. Chandrasekaran and his team have developed an algorithm that can recognize wit and can even construct some by themselves. Regardless of the absence of any context and knowledge of the diverse human culture, this algorithm lets machines predict whether a deed is witty or not.
To start with the research, a limited set of pictures were created using clip art. These clip arts contained human and animal models depicting different emotions which were placed around different objects such as tables, curtains, etc. Post this, the pictures were judged by humans and based on their judgments of how an image/object contribute to humor, a related algorithm was developed, capable of understanding the difference between humor and distinctive emotions.
Upon testing it was seen that the algorithm could transform funny images to unwitty ones by a large percentage. This transformation was achieved by replacing certain living objects with non-living ones. Though the algorithm did not perform well in making images more amusing, it still improves the scenes with a 28% success rate.
Humorous scenes on the left and their corresponding humorless counterparts on the right were examined by the Artificial Intelligence algorithm.
In the past, too many algorithms were developed, like a joke writing machine known as Manatee and SASI which was a sarcasm detector. But none understood the most basic and worthwhile human expression-Humor.
With the development of visual humor algorithms, researchers are hopeful that it will contribute in many applications like picture editing tools, smart world cameras can capture that exact moment of bliss. These humors detecting tools can surely bring a difference to the social media as well as to the marketing world.