Scientists from the University of Waterloo in Canada have done the impossible. They've created a functioning model of the human brain, a complex simulation of millions of interconnected neurons firing bolts of electricity back and forth and completing complex tasks. It works. It just doesn't work nearly as well as a real human brain. In fact, it's not even close.The new computer brain is called Spaun, and it's a breakthrough to be sure. An acronym for the Semantic Pointer Architecture Unified Network, Spaun is perhaps the most capable simulated brain that scientists have ever build, and they did it using fewer neurons than past attempts. Spaun contains 2.5 million virtual neurons, far less than the billion or so neurons working in IBM's cognitive computing project and just a tiny fraction of the 86 billion neurons in the average human brain.And while our big brains can dream up crazy things like like _The Canterbury Tales_, space ships and, well, simulated computer brains, it's a major accomplishment for Spaun to perform even the simplest of tasks. Counting, for instance. Not only can Spaun count, it can store and recall memories and even perform basic intelligence tasks like guessing the next digit in a series of random numbers.Don't go thinking this new brain simulation is going to take over the world anytime soon, though. Its limited neural network can't learn new tasks other than the ones that head researcher Chris Eliasmith, a theoretical neuroscientist, programmed into it, and its comprehension of input is limited to ten numbers and a few symbols. Spaun is also quite slow. Tasks that would take just one second for a human brain to perform can take hours for Spaun to simulate.As brain experts are quick to argue, the point of brain simulations is not to create artificial intelligence or super powerful computing capabilities. We have artificially intelligent software and supercomputers for those kinds of things. Spaun is a breakthrough because it appears to behave more like the brain than any previous simulations."Until now, the race was who could get a human-sized brain simulation running, regardless of what behaviours and functions such simulation exhibits," Eugene Izhikevich, chairman of the Brain Corporation, explained to Nature. "From now on, the race is more [about] who can get the most biological functions and animal-like behaviors. So far, Spaun is the winner." And if anybody ought to be the judge of the best fake brain, it ought to be the chairman of the Brain Corporation.Ultimately, better simulations like Spaun will give neuroscientists a better idea of what's going on when human brains start to slip. "If we destroy parts of this model, we can see what behaviors might fail," Eliasmith explained. "Or we could change how neurotransmitters function and see how that relates to behavior."For example, the research team has already successfully reproduced the neural decay of aging by killing off Spaun's neurons at the same rate that neurons die when we age. Sure enough, the brain showed the same signs of dementia. Now, we just need to teach Spaun a bunch of stories about life during the Depression, the invention of TV and the problems with Communism, and it'll be just like your awesome grandpa.
Advertisement