Taking AI at its word
Harvard University and the MIT-IBM Watson AI Lab have developed a tool to identify text generated using artificial intelligence. The Giant Language Model Test Room determines human authenticity by looking for statistical patterns that are apparent when AI is at work.
“We make the assumption that computer-generated text fools humans by sticking to the most likely words at each position, a trick that fools humans,” the researchers say. “In contrast, natural writing actually more frequently selects unpredictable words that make sense to the domain. That means that we can detect whether a text actually looks too likely to be from a human writer.”