An article published by Nature on February 23rd highlights Henry Markram’s proposal to spend €1,ooo,ooo,ooo to create a supercomputer simulation “that integrates everything known about the human brain, from the structures of ion channels in neural cell membranes up to mechanisms behind conscious decision-making.”
To say that Markram’s idea is controversial is to state the very obvious.
Who is this guy, and why does he want to corner most of the available research money for a project that many believe will spend tons of effort and cash on what may be the wrong approach to brain science in the first place?
Markram, a brain electrophysiologist, is trying to duplicate with his Human Brain Project (HBP) what the Human Genome Project did for genetics. Markram argues:
“Brain researchers are generating 60,000 papers per year. They’re all beautiful, fantastic studies — but all focused on their one little corner: this molecule, this brain region, this function, this map.”
The Nature article reports: “With computers, Markram realized, you could encode all of those models explicitly and get them to work together. That would help researchers to find the gaps and contradictions in their knowledge and identify the experiments needed to resolve them.”
Markram doesn’t want just to model neurons as points on a grid. He wants to model the functioning of the brain completely: “from the genetic level, the molecular level, the neurons and synapses, how microcircuits are formed, macrocircuits, mesocircuits, brain areas — until we get to understand how to link these levels, all the way up to behaviour and cognition.”
Markram’s project would require, among other things, enormous computing power. He had little prospect of finding the massive grant money he needed until the European Union announced two years ago that it was prepared to allocate €1 billion each to two “high-risk, but potentially transformational” projects. In May 2011, the HBP was one of six candidates that received seed money to prepare a full proposal, due in May 2012.
Many other neuroscientists are skeptical that Markram’s commitment to extreme detail will clarify, rather than muddle, our understanding of the brain. And they question whether Markram, whose work so far has concentrated on a small series of ion pathways in rats, can build the model at all. “The tiny swathe of simulated rat cortex has no inputs from sensory organs or outputs to other parts of the brain, and produces almost no interesting behaviour, pointed out Kevan Martin, co-director of the [Institute for Neuroinformatics].”
The question most relevant to the recent focus of this blog is, of course, If Markram does achieve his goal, what will he have? A very expensive database? The missing tool for a real understanding of human consciousness? A start to the end of neurological diseases? A self-aggrandizing boondoggle?
Many critics of the Human Genome Project thought that it was an over-hyped, pointless, and extremely expensive exercise in sterile description. A card catalogue (yes, I’m showing my age here) is not the books it describes. Having a complete listing of the holdings of your local library doesn’t make you well-read. That sort of criticism.
Yet, in the last decade, how much new knowledge, how many practical applications, have risen from the bare “catalogue” of human genetics?
Perhaps, the same kind of thing will happen if Markram’s model is constructed. Perhaps, if you build it, knowledge will come. That’s what its advocates hope, certainly, although I suspect that if the project is completed any new insights will come in unexpected and surprising ways.
That’s the way that these things go, isn’t it?