Making Progress With Robots Understanding Visual Humor

Robots Make Inroads on Understanding Visual Humor

Aaron Krumins | January 20, 2016

You might think something as abstract as humor would be the piece de resistance for robots in terms of mimicking human abilities. Not so: Researchers toiling at Virginia Tech University have created a machine learning algorithm that can alter a visual image to make it funnier with a 28% success rate.
That result compares rather favorably with my own joke telling ability, a fact that was impressed upon me during last month’s yuletide family gathering in which this journalist’s attempts at brevity went over like a lead balloon. If the Virginia Tech research is anything to go by, in the future we may look forward to outsourcing the witty interjections at our Christmas gatherings to the domestic helper bot – a prospect that will likely scare or excite, depending where you fall on the Luddite spectrum.

Artificial intelligence algortihm taught to recognise humor. Source: http://www.newsweek.com/artificial-intelligence-algorithm-taught-recognise-humor-413832

<more at http://www.extremetech.com/extreme/221383-robots-make-inroads-on-understanding-visual-humor; related links: http://www.newsweek.com/artificial-intelligence-algorithm-taught-recognise-humor-413832 (Artificial Intelligence Algorithm Taught to Recognise Humore. January 11, 2016) and http://www.technologyreview.com/view/545316/ai-algorithm-identifies-humorous-pictures/ (AI Algorithm Identifies Humorous Pictures. The latest work with AI machines is expanding the new field of computational humor. January 8, 2016)>