Humans Are Actually Not That Stupid

This article can be found at http://www.nytimes.com/2010/10/05/science/05compute.html?_r=1&ref=science

    Like stem cell research, artificial intelligence is a field of study where researchers proceed with caution. There are potential consequences that could overshadow the benefits—would machines that are as intelligent as humans be the best thing to have around? It is both a practical and a moral question. The existence of intelligent robots would basically mean we created a whole new race of man, or perhaps a whole new species of being. Can we do that? Should we do that?
    Well scientists are trying to answer the former before they get to the latter. Machines are a long ways away from exhibiting human intelligence, but artificial intelligence research is in full throttle across the globe, as scientists plug away in the effort to bring the future closer. This story is about a huge stepping stone in that direction—the Never-Ending Language Learning system (NELL). This mega computer browses the Web and essentially teaches itself (with some help from scientists of course). A great example of how far computers have to go is the fact that NELL would not be able to understand the sentences, “The girl caught the butterfly with the spots,” and “The girl caught the butterfly with the net.” We immediately understand that the butterfly has the spots, not the girl, and the girl has the net, not the butterfly. This comes from our cumulative knowledge, building and evolving over years and years of experience. NELL has none of this. Scientists hope that through its ability to search the Web and acquire its own cumulative knowledge, it will someday have human-like language understanding.
    The story is very well-written. Starting with the current gap between computers and humans, then moving in to the details of NELL, all while expanding on the topics of semantic computing and artificial intelligence, the writer Steve Lohr does a brilliant job drawing the reader in and keeping him/her engaged. The examples he chooses of NELL’s abilities and inabilities, as well as the general abilities/inabilities of today’s computing, work spectacularly.