First it was about machines which would work for us. Now the venerable white coats are talking about machines that will crack jokes for us. At the risk of sounding like downright Luddites shaking our pitchforks, may we simply sound out that old warning: the end of the world is nigh. Humourless researchers at the University of Cincinnati have now reportedly equipped computers with a sense of humour. They obviously don’t want to tell the world that they have spent valuable grant money on teaching personal computers to ask you, “What’s the difference between blondes and traffic signs?” The researchers are apparently developing software that can solve problems that are “informally stated” and can, therefore, interact with us humans more naturally. That, we have been told, includes the ability to crack jokes.
Cincinnati 2007 is a far cry from Stanley Kubrick-style 2001 Outer Space (for one lunar travel is not that easy in the real world yet). But may we gently remind the scientists at the university’s Applied Artificial Intelligence Laboratory what happened to HAL-9000, that computer in the film, 2001: A Space Odyssey. It starts getting informal — far too informal for comfort — and turns into a rogue homicidal machine. The scientists obviously are pointing to the direction of making machines think like humans. For that purpose, the computer will include a knowledge base culled from a children’s dictionary and children’s text (easy words) and a special algorithm, or programming equation, that will decipher what words sound like, how they’re spelt and what they mean. And familiarising jokes — puns included — is the easiest way to go about such things.
Which brings us to the question: will these smart PCs only make PC jokes? Or will they crash once they know that the difference between blondes and traffic signals is that the latter can say stop.