Friday, March 30, 2012


The brain is flesh. A softball-sized clump of meat. Blood and burger.

But, oh, what meat!

An estimated 100 billion neurons. Each neuron reaching out with spidery arms to touch thousands of others. Or almost touch, at connections called synapses. Each synapse in any one of about ten levels of excitation. One hundred trillion or so synapses, spidery tendrils almost touching.

The mind, we'll agree, is inseparable from the brain. Or perhaps we should say that the mind is inseparable from the body, with its various instruments of perception. Or better: The mind is inseparable from the brain in interaction with the world.

In any case, we won't understand the mind -- consciousness, self-awareness, memory, reflective thought -- until we understand the brain.

Brain researchers are generating 60,000 papers a year, according to a report in the 23 February issue of Nature. Most focus on some small aspect of the brain -- this region, that molecule, these synapses, and so on. Henry Markram, a south-African-born electrophysiologist at the Swiss Federal Institute of Technology in Lausanne, wants to build a supercomputer that integrates everything known about the brain, something he calls the Human Brain Project (HBP). His project is one of six finalists vying to win 1 billion euros as one of the European Union's two new decade-long "Flagship initiatives."

A billion euros to build a box as complicated as the human brain.

The computer power required to run such a complex model would be roughly an exaflop, or 10 to the 18th operations per second. This technological milestone will probably be met by the 2020s. In the meantime, Markram wants to get started. He has put together an international team of investigators. They hope to have achieved a functional model of a rat's brain by 2014.

This is like the human genome project writ large.

Will a machine that mimics the hundred trillion synaptic connections of the human brain be conscious? Self-aware? Formulate language? Have "moments of being"? Or is there something at work in consciousness that eludes fundamental science as we understand it today? These are big, big questions, perhaps the biggest questions the human mind can ask and potentially answer.

Not all brain researchers are as cheerfully optimistic as Markram, and worry that HBP will suck money out of other, more promising research. Markram sees HBP as a way of integrating research that presently proceeds in relative isolation.