Manuel Blum, Carnegie Mellon University
Luis von Ahn, Carnegie Mellon University
The researchers are investigating several novel techniques for utilizing the computational abilities, or “cycles,” of humans to solve problems that computers cannot yet solve. For example, tasks such as image recognition are easily performed by humans, whereas they continue to challenge the most sophisticated computer programs. Where traditional approaches to solving such problems focus on finding and improving various computational techniques, these researchers are using a novel approach: constructively channel humans to do the work. This approach has a vast number of applications in areas as diverse as security, computer vision, Internet accessibility, adult-content filtering, and Internet search.
This groundbreaking research is exemplified by the researchers’ previous work on the ESP Game. This is a seductive, collaborative online game: many people play over 40 hours a week and, in the process, provide meaningful, accurate labels for images on the Web. Among other things, these labels can be used to improve image search accuracy and to help block undesirable content. In a few months, the ESP Game has collected over 8 million image labels and, if deployed at a popular gaming site, could label all images on the Web in a matter of weeks.