IBM has developed a microprocessor capable of figuratively “rewiring” its connections as it encounters new information, similar to how biological synapses work, thereby emulating the human brain. That is how the system could learn, rather than be programmed.
Synaptic connections between brain cells physically connect themselves depending on experiences. The process of learning is essentially the forming and strengthening of these connections. A computer can simulate such a system by paying more “attention” to important inputs, and paying less attention to others.
Also, the human brain can perform complex tasks rapidly and accurately using “as much energy as a 20 watt light bulb in a space equivalent to a 2 liter soda bottle”.
Researchers at IBM have been working on a cognitive computing project called Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE), led by team leader Dharmendra Modha, where the eventual goal is to develop a computer that is of the efficiency, size and powr of the human brain. So far the computer has been taught to recognize handwriting, play Pong, and guide a car around a track. Apart from a vast number of learning tasks, like face recognition, behavioral predictions and ecosystem models, the future of cognitive computing could reach a point where machine consciousness is possible.
The concept is larger than life, or more promisingly, just as real as life.