Recently, [url=http://www.eurekalert.org/pub_releases/2011-04/uosc-rcf042111.php]resear... at the University of Southern California created an artificial neuron.[/url] Sure, it's simplified, and it's only one, but it's a 5 micrometre carbon nanotube chip that simulates a neuron. Not something to be sniffed at. Now, something that's come to my mind a lot lately is the necessity of a cross-over in the neuroscience and AI research fields. If we want to greatly improve the latter, we need to look at the former. Things like the ability to learn and later access/recall information are crucial if we ever want to create an AGI and every biological organism on Earth with a brain can do this to at least a rudimentary level. So my question is, given the level of funding that CERN has put towards building the Large Hadron Collider, what would the possibility be of a large, dedicated AI research group gathering an equivalent amount of funding and using it to build a gigantic supercomputer, so large and powerful as to allow extremely high fidelity simulation of neurons, to allow a brute-force realtime simulation of a brain? If we must, could we go so far as to simply create a physical architecture that functions like a brain and simply reverse-engineer it in situ? We could advance both fields, neuroscience and AI research alike, at an astonishing pace. It'd be a hell of a lot of work, and extremely expensive, but the benefits would be worth it and see far faster return on investments than most any other scientific project of that scale. If it's possible, that is. Thoughts?