according to spike activity at mindhacks, the human brain project
…sells itself as aiming to “simulate a complete human brain in a supercomputer” but this is clearly bollocks.
It’s interesting that this claim makes the press kit and the flashy video but the actual report (pdf) has much more sober claims about ‘simulating brain dynamics’ and the like.
But it’s important to realise that while their big sell is nonsense, the project is likely to genuinely revolutionise neuroscience in a way that could push the field light years ahead.
What Markram has realised is that the single biggest barrier to progress in neuroscience is the co-ordination, sharing and integration of data.
Essentially, it’s a problem of information architecture but quite frankly, you can’t sell that to politicians and they can’t sell it to the public. Hence the ‘simulating a complete human brain’ fluff.
What the project aims to do is co-ordinate neuroscience teams looking at neurobiology, cognitive neuroscience and computational modelling and give them the tools to easily share data with each other.
One of the big pay-offs will genuinely be the creation of biologically feasible computer simulations of neural networks with the hope that these can be used for practical applications like virtual drug testing and computer-based experiments.
Markram has gained valuable experience of meshing heavy-duty computing with working lab teams and has recruited some of the world’s leading neuroscientists to the project.
Although the spin seems over-the-top scientifically this is an important project that, if successful, could be a scientific landmark.







