The Quantized Mind, part 2: Connectomics and Blue Brain
Hi, guys. A few weeks ago I wrote the first post in a series that I am calling “The Quantized Mind.” You can read part 1 under the February 23rd, 2009 entry on this blog. In part 1, I lay out an argument that there will eventually be an opportunity for human beings to measure phenomena such as cognition in quantifiable terms. Making the argument is the easy part. Making it happen is where the heavy lifting begins.
In this post, I want to introduce the concept of connectomics. Like all body organs, the brain is made up of individual cells. This realization that the brain was composed of cells was actually a radical breakthrough that occurred early in the 20th century. The neuroscientist Santiago Ramón y Cajal deserves a generous portion of the credit for advancing our current understanding of the cellular components in the brain for his outstanding work in characterizing neurons and their types.
The number of brain cells, i.e., neurons, that make up the brain is absolutely staggering. It is estimated that there are 10 billion brain cells (10,000,000,000), each of which has thousands of connections to other brain cells. To do a mental experiment with this number, imagine that the each person in the entire population of China has a thousand pieces of string in his or her hands. Now imagine that every person holds on to one end of each piece of string, and then gives the other end of each of the thousand pieces of string to a thousand other individual people. The massive web of people and pieces of string is orders and orders and orders of magnitude beyond what our human minds can really imagine. Now take this unimaginably huge and complex web of people and string, multiply it by about ten, and you will have a conservative estimate for how complex the human brain is.
When we talk about quantifying phenomena such as cognition, we need to employ some realistic humility to acknowledge that we are itsby-bitsy teeny-weeny neonates in understanding our own brains. Winfried Denk, a neuroscientist at the Max Planck institute, has estimated that it would take three billion person years to map out just one narrow functional unit in the brain known as a cortical column—to say nothing of completing the project to map out the brain in its entirety. Hardly something to hold our breath for!
The good news, though, is that it is not just one person doing all the work, and that it’s certainly not just humans doing all the work, either. We are benefiting from exponentially increasing capabilities to resolve these problems computationally, both in terms of the exponentially increasing computer power at our disposal, and also in terms of the exponentially increasing understanding of brain connection dynamics (i.e., the more we understand, the more that this understanding allows us to understand). Predicting timescales for the convergent effects of multiple exponential factors is tricky, to say the very least, and I don’t know of any good paradigms that convincingly argue an estimated time of arrival for our ultimate understanding of brain wiring. Ray Kurzweil, the famous and infamous technology forecaster, has made bold prediction that humanity will have successfully reverse-engineered the human brain by the 2040s. Don’t get me wrong—I think that would be amazing. 30 more years is a heckuvalot shorter than three billion person years hacking away at cortical columns. But I guess the best that I can say is let’s see what happens.
Probably one of the most serious efforts to rigorously decode the mystery of the human brain is the Blue Brain Project at École Polytechnique (EPFL) in Lausanne, Switzerland. In 2005, the project team successfully completed a ten year endeavor to map the cortical column of a rat. As reported in the June 2005 New Scientist, the current goals of the Blue Brain Project are:
- construction of a simulation on the molecular level, which is desirable since it allows to study effects of gene expression;
- simplification of the column simulation to allow for parallel simulation of large numbers of connected columns, with the ultimate goal of simulating a whole neocortex.
Before we get too cozy in our seats, let’s consider that the whole human neocortex is estimated to contain 1 million cortical columns. In other words, once we complete the first goal to construct a molecular simulation of a single cortical column, it’s not exactly a hop, skip, and a jump to simulating our entire brain. Nonetheless, we have many factors working in our favor: not only are our computer technologies rapidly augmenting our power to handle some of these computational problems, but the general public interest in our brain is a very good thing for the work of neuroscientists. Strong international cultural leaders such as the Dalai Lama are openly publicizing the humanitarian good that could potentially come from better understanding our inner workings in high-resolution detail. Combined with recent sociopolitical forces to renew the emphasis on science research and funding, there are strong reasons for optimism about the path that lies ahead.