According to a an article from Fudzilla, scientists at Carnegie Mellon University have built a software that can search the web on a non-stop basis and learn common sense.
A normal human that browses the internet daily can learn a handful of information. However the software, dubbed the Never Ending Image Learner (NEIL), was designed to search for images and do its best to understand these images on its own. The program runs on two clusters of computers that include 200 processing cores, way beyond the ability to process information for a normal individual. As NEIL grows a visual database it is expected to gather common sense on what is being called as “massive scale”.
The designers have already shown some unique findings that could relate to common sense, such as “Deer can be a kind of / look similar to Antelope,” and “Trading Floor can be / can have Crowded”. The results so far are not really spectacular, but it is sign of progress. Computers do not have the ability to comprehend, that’s why software is based on conditions and functions.
Abhinav Gupta, assistant research professor in Carnegie Mellon’s Robotics Institute said that images were the best way to learn visual properties. People learn this by themselves and, with NEIL, computers could gain that ability as well.