Microsoft Word Kurzweil, Ray The Singularity Is Near doc

Brain Complexity.
Thomas Ray also makes the point that we might have difficulty creating a system equivalent to 
"billions of lines of code," which is the level of complexity he attributes to the human brain. This figure, however, is 
highly inflated, for as we have seen our brains are created from a genome of only about thirty to one hundred million 
bytes of unique information (eight hundred million bytes without compression, but compression is clearly feasible 
given the massive redundancy), of which perhaps two thirds describe the principles of operation of the brain. It is self-
organizing processes that incorporate significant elements of randomness (as well as exposure to the real world) that 
enable so relatively small an amount of design information to be expanded to the thousands of trillions of bytes of 
information represented in a mature human brain. Similarly, the task of creating human-level intelligence in a 
nonbiological entity will involve creating not a massive expert system comprising billions of rules or lines of code but 
rather a learning, chaotic, self-organizing system, one that is ultimately biologically inspired. 
Ray goes on to write, "The engineers among us might propose nanomolecular devices with fullerene switches, or 
even DNA-like computers. But I am sure they would never think of neurons. Neurons are astronomically large 
structures compared to the molecules we are starting with." 
This is exactly my own point. The purpose of reverse engineering the human brain is not to copy the digestive or 
other unwieldy processes of biological neurons but rather to understand their key information-processing methods. The 
feasibility of doing this has already been demonstrated in dozens of contemporary projects. The complexity of the 
neuron clusters being emulated is scaling up by orders of magnitude, along with all of our other technological 
capabilities.