Tuesday, May 25, 2010

Our Brain's Operating System

This is a neat description of the architectural differences between the organic brain and the digital computer that we know so well.  This is a case of asking an interesting question driven by our experience with hardware.  Why does ours work so well?
It is worth the read for that insight.  I am sure this will find itself into textbooks.
 The interesting question is whether we could use these strategic insights to produce a new operating system framework.  I would like to see a development team try but few are likely to see much profit when we are talking about completely new code all the way through to get it all right.
So we will continue to practice crash and burn.
Wonder why we don't crash like computers? Yale explains
16:55 May 11, 2010

the control network of bacterium E. Coli. Left, and the Linux operating system, right

Whether right or for wrong, the human brain is often compared to a computer, and vice-versa. They both receive data, process it, store it, and output new data. Unlike computers, however, the human brain doesn’t crash. Yes, people have nervous breakdowns, but that has more to do with psychological stress than with data management. Now, researchers from Yale University have figured out why our brains succeed where computers fail.

The research team compared the genome of E coli bacteria with the Linux operating system. Both of the control networks, it turns out, are arranged in hierarchies. In E coli, the molecular networks are arranged in a pyramid. A limited number of master regulatory genes sit at the top, controlling a wide range of specialized functions beneath them.

By contrast, Linux is more like an inverted pyramid - numerous routines are at the top, controlling a few generic functions at the bottom. This is because software engineers save time and money by building on existing routines, instead of starting systems from scratch. Such an approach makes the system vulnerable to breakdowns, however, as even simple changes to a generic routine can be very disruptive. To minimize problems, the generic components need to be continually fine-tuned by software designers.

The Yale scientists noted that in a living organism, generic components that need to be constantly updated would not be a good survival trait. Instead, over billions of years of evolution, the E coli bacteria has evolved many highly specialized modules. Together, these modules are ready to handle most eventualities, resulting in a much more robust network.

No comments: