Artificial intelligence doesn’t come cheap. Researchers from IBM and five universities are getting $4.9 million in funding from the U.S. Defense Advanced Research Projects Agency (DARPA), to develop computing systems that they say will emulate the brain’s abilities for sensation, perception, action, interaction and cognition.
Dharmendra Modha, manager of IBM’s cognitive computing initiative, is leading the effort. He says the mind is more powerful and nimble than the world’s fastest computers (yet is so small and cheaply produced, rather like a Honda). His goal is to develop computing systems with brain-like abilities to process massive amounts of data from a variety of sources, while using nanotechnology to recreate the brain’s small-scale synapses and neurons.
Such a machine would be able to handle things that we find rather emotionally messy, like ambiguity. And with that processing power like that, it could watch all our TV shows for us.
Obviously, we’re not likely to see a human brain-like computer anytime soon, unless we’re watching a re-run of “Rain Man.” The researchers - which include folks from Stanford, University of Wisconsin-Madison, Cornell, Columbia and University of California-Merced - will give a progress report in nine months.
Trackback
One comment so far
Neural Networks do simulate human intelligence. They store the knowledge in the weights. We could set a neural network to an advanced state of training by storing the weights learnt so far and train it further.
With respect to the human brain we do not know of any such technique of transfering knowledge to the brain directly. We hope the researchers would think of transfering knowledge directly in addition to teaching the brain-like computers.
- Posted by Mohideen Ibramsha