A programme written by researchers at the University of Edinburgh came up with one liners such as “I like my men like I like my monoxide – odourless” although some of its less successful attempts suggest artificial intelligence is a long way off.
The software searches huge amounts of data to create jokes with unlikely pairings of words that have an unusual connection between them.
The jokes made study volunteers laugh, although not as much as human-created gags gathered from Twitter.
David Matthews [no joke, given the reporter’s name – ed], a research postgraduate at Edinburgh’s School of Informatics, explained that computers have an advantage over humans in creating these kinds of jokes because they can process so much information.
He added that the “holy grail for machine-generated comedy” would be to also include cultural references, “but these are very hard to capture,” he said.
The study should help computers better understand and process language, the researchers said.
The computer gags, which range from the inventive to the bafflingly surreal, include:
- I like my coffee like I like my war – cold
- I like my relationships like I like my source – open
- I like my men like I like my acorns, buried
- I like my women like I like my gas – natural
- I like my fish like I like my text - raw
- I like my men like I like my court, superior
- I like my business like I like my fish, small
- I like my boys like I like my sectors – bad
- I like my men like I like my monoxide – odourless
Its findings are being presented next week at the Association for Computational Linguistics annual meeting in Sofia, Bulgaria.
News of the research comes as thousands of people gather for the start of the UK’s biggest annual platform for comedy, the Edinburgh Festival Fringe.