RESEARCH:
The exhibition is distinguished by its focus on the research and experimentation of artificial intelligence as much as the excitement of the final product. The moment visitors enter the exhibition, they are participating in research at MIT. Many of the exhibits are both experimental and experiential, with the visitor getting a behind-the-scenes look at the process of evolution that precedes a successful invention.
On the introductory panel to Robots and Beyond, Professor Rodney A. Brooks, director of the Artificial Intelligence Laboratory, notes, "Artificial intelligence at MIT has always been about thinking large and implementing daring ideas. At the core, we want to understand how to make intelligent machines and how it is that humans operate. We build machines that perceive, understand language, have common sense, learn, and act in the world."
The exhibition details the research behind building an intelligent robot at MIT – one that interacts with its environment in human-like ways.
HOW DOES A ROBOT LEARN?
Cog, a famous humanoid MIT robot, currently part of a traveling robotics exhibition, was designed to translate its environment through tactile and visual sensors. Through prototypes and other media, the visitor can follow the evolution of Cog from a fairly primitive robot to one of sensory sophistication.
"The motivation behind Cog," states Brooks, "is the hypothesis that humanoid intelligence requires humanoid interactions with the world."
Another renowned anthropomorphic robot also on exhibit, Kismet, was built to communicate its needs and wants to people using human-like facial expressions, body position, gaze direction, and voice. Visitors can view a series of photos that capture the full range of Kismet's expressions.
ARE ROBOTS SMARTER THAN HUMANS?
Robots and Beyond illustrates what people learn from computers and what computers learn from people. In many ways, artificial intelligence is infinitely less intelligent than human beings. A two-year-old child, for example, can instantly tell the difference between a dog and a cat, but a computer would find the differentiation difficult.
Both cat and dog have four legs and a tail, some dogs are as small as cats, so a computer could mistake a small dog for a cat. Not all barks and meows are standard either, so sound does not necessarily clear up the computer's confusion. Similarly, a robot has a difficult time distinguishing a man from a woman—there are simply too many superficial variables.
But computers leave human beings in the dust in other areas. They can discern intricate patterns and identify complex matching patterns, such as in fingerprints. Human beings would not have the ability to read each other's fingerprints or irises to confirm identification, but computers can instantly process the complex network of patterns and determine if one matches another.