“I have a friend of mine who’s put together a project to put an AI brain into a lifelike robot, the goal being to make it as smart as a stupid three-year-old, from which point it can begin to learn on it’s own and develop consciousness.” The robot is being developed by a team of roboticist rock stars, including David Hanson (who built the Philip K. Dick android head) and Mark Tilden (creator of BEAM robotics and the WowWee Robosapien robot). Help Make Me the World’s Smartest Robot
Though I’m sure this project will yield some impressive results, I highly doubt that the result can be classified as consciousness, or even ‘common sense’. The robot will probably do a very good job at simulating consciousness, but will not be imbued with it.
Personally, I put my faith in the work of Steve Grand who, for many years now, is building virtual brains via a bottom-up approach (in contrast to the modular, top-down approach described above), neuron by neuron.
Steve’s approach appeals to me so much because his model doesn’t simulate behaviour, but it simulates neurons. The behaviour this type of brain will display will be real, even though the brain itself isn’t.
Anyone interested in this subject matter and who is not yet familiar with Steve’s work, should definitely read his books.
http://www.amazon.co.uk/Growing-Lucy-Build-Android-Twenty/dp/0753818051/ref=pd_sim_sbs_b_1