The creation of Artificial Intelligence–computers that talk like real human beings–has been extremely challenging. Significant progress has been made, but teaching computers to recognize things like sarcasm has been notably difficult. This week, researchers at Cornell University made big news when they allowed to computers called “Cleverbots” (AI programs designed to interact with people in chatrooms) to interact with each other. The result has led some to speculate that the AI machines have been taking their cues from hostile, LSD-tripping chatroom participants. Folks who were already worried by HAL (the famous AI computer from Clarke’s Space Oddessy) are probably wracked with a new list of concerns.
You can watch the two Cleverbots fighting with one another in this link. The conversation includes such memorable lines as “I am not a robot. I am a unicorn.” And, “You are mistaken. Which is odd, since memory shouldn’t be a problem for you.” Maybe the next challenge in AI ought to include programming tutorials on politeness.