Humans are very good at recognizing patterns. We even think we find patterns in places where no pattern really exists. Some good examples are the constellations. An astronomical constellation is a manmade concept. The different points of light that make up Orion or the Big Dipper bear no natural relation to one another -- in fact, the stars that make up these constellations are billions of light years apart. Yet when we look up at the night sky, our minds begin to group stars into patterns and shapes. It is this pattern-seeking behavior and imaginative vision that allows us to recognize people and objects.
Computers have trouble recognizing some types of patterns. If the pattern is easily reduced to numerical data, like sports statistics or even a DNA strand, a well-written computer program will usually have no trouble finding it. However, patterns that can't easily be described with numbers are all but invisible to the same software. If given enough sample data and careful instruction, for example, a digital camera might be able to automatically distinguish the image of a human face from that of a tree branch or a basketball. However, ask a modern digital camera to describe the mood of the person in the picture, and you're likely to see one skill set in which humans currently hold the advantage.
This is not to say that computers will never be able to analyze a human face and read its emotions the way you can. For you, this analysis is a simple, intuitive move: You look at the face and use your intuition to instantly recognize emotional content. For the computer, the evaluation requires painstaking entry and analysis of face image data before it can begin to correlate face images to emotional values. This discrepancy shows, in a way, how incredibly complex the human brain really is. So much so that we still don't entirely understand how it works. It's hard for us to build a simulation of something we don't yet fully grasp ourselves. So far, computers still require pre-programmed instructions, and they know what they're told to know. When the time comes that they can adapt to new situations and process problems without being programmed to do so, then they will be truly intelligent and not just sophisticated calculators. Futurist Dr. Ray Kurzweil predicts that computers will one day be self-aware, analytical and able to adapt themselves in order to improve their performance. In other words, they will exhibit recursive self-improvement. After this, it might be hard to imagine any task that a computer will not be able to perform better than a human.
What is the Semantic Web?
Answered by HowStuffWorks
What is the Semantic Web's technical foundation?
Answered by Science Channel
Is Watson a thinking machine?
Answered by W. Daniel Hillis