Computer-related regrets are an everyday occurrence, whether they involve accidentally hitting the Reply All button, instant messaging the very person you want to complain about or being tagged in a drunken Facebook photo. But what if your computer could learn to regret mistakes just like you do?
Google is funding a project at Tel Aviv University that plans to do just that.
Yishay Mansour of Tel Aviv University's Blavatnik School of Computer Science is spearheading an effort to help computers better recognize the distance between a desired outcome and the actual outcome. In a sense, this is regret. But it doesn't stop there; the program aims to help computers learn from those mistakes so they can more accurately predict events in the future, thus avoiding more "regret."
Google is interested in this project because it has the potential for algorithms that can more efficiently route traffic and avoid overload from traffic spikes. Specifically, Google anticipates it could improve its AdWords and Adsense performance.
"If the servers and routing systems of the Internet could see and evaluate all the relevant variables in advance, they could more efficiently prioritize server resource requests, load documents and route visitors to an Internet site, for instance," Mansour said in a statement. "We are able to change and influence the decision-making of computers in real-time. Compared to human beings, help systems can much more quickly process all the available information to estimate the future as events unfold — whether it's a bidding war on an online auction site, a sudden spike of traffic to a media website or demand for an online product."
Humans aren't predictable individuals, but the algorithm can process and predict trends of large groups of users, and it learns as it goes so that over time it learns from mistakes and becomes better.
Of course, this project won't make computers actually feel the emotion of regret, but it will make them better able to respond to mistakes and unforeseen consequences. Such a development has huge implications — not just for Google's servers, but for artificial intelligence research and future consumer products.