Remembering Aaron Swartz - Raw Thought

Be curious; Read widely. Try new things. I think a lot of what people call intelligence just boils down to curiosity.

 - Aaron Swartz

Aaron Swartz (1986 - 2013)

Last week, I was saddened to hear the news about the death of Aaron Swartz. I have never met Aaron in person but I was a regular follower of his blog and work for many years. Both of these led me to deeply admire and respect him. His tireless work against the passing of the SOPA bill was of significant interest to me because of the serious implications of the bill. As a programmer his talents were legendary and as an activist his tireless efforts admirable. His passing is indeed a great loss. The Economist has a nice obituary here that perfectly reflects this sentiment. 

As an avid reader of his blog, I am deeply saddened to know that there will not be anymore updates to the site. So I decided to collect all of his blog posts over the years and compile it into a PDF/Ebook. So this past week, I wrote a simple Python script that crawled Aaron's weblog and retrieved all of the posts one by one and complied them into a single file.

Human Brain vs the Super Computer ?

This post is cross-posted from http://goo.gl/7vsEk. Pardon the grammatical errors, but I posted this as a reference for myself.




The modern science believe that if the human brain can work to its full potential, free of all the fears, tension, complexes etc., it can be much more efficient (2-3 times) than the super computer.

The major functions of the super computer are memory and speed. Memory is confined in the Bites and speed in FLOPS (Floating Point Operations per Second). One thousand Bytes are defined by K and one thousand K from a Mega. The memory of a standard PC runs from 16 MB (Mega Bytes) to 22 MB. The storage space (of data) though goes upto many Giga Bytes (1000 Mega=1 Giga). In spool type storage computers, the Bytes are calculated in Tera Bytes (100 Giga=1Tera). On the other hand, human brain uses Neurons (or brain cells) to store the memory. Although the scientists have not been able to put together a definite figure, recent findings put it to approx. 20 billion. These Neurons are, in turn, connected with each other through fibrous tissues. These joints are virtually uncountable and the approx. figure is put at 1000 billion. If we consider each joint as one byte, the approx. capacity would be 1000 Giga Bytes or 10 Tera Byte. As a RAM, this is much more than any super computer.
As far as the speed is concerned, speed of a modern computer is calculated in FLOPS or Floating Point Operations per Second. K2 super computer's speed is approx. 1 Giga FLOPS. K2 was one of the earlier computers. By 1990 the speed was improvised ten times; being 10 Giga FLOPS. Further innovation came in CM5 in 1992 which ran at 100 Giga FLOPS. The latest one, however, has surpassed all earlier inventions and has a speed of 2 Tera FLOPS. We shall analyze the speed of human brain in comparison.

If there is an electrical vibration in Neuron, it is considered activated. When the electric charge flows in the Neuron, it is said that the Neuron is "Fired". At a time 1% to 10% Neurons are fired in the brain. Each Neuron, in turn, fires (or vibrates) 100 times per second. From the total Neurons of 20 billion, if 1% to 10%, i.e. 200 million to 2 billion Neurons fire 100 times per second, we can surmise the speed of the brain. If one Fire of Neuron is considered 1 FLOPS, the speed (of only 1% to 10% capacity) would be from 20 giga FLOPS to 200 giga FLOPS. If we assume that all the Neurons Fire in the brain at one time at a rate of 100 Fires per second, the total speed would be 20 billion x 100 FLOPS per second; i.e. 2000 giga or 20 Tera FLOPS that is much beyond the speed of any super computer.

We can, therefore, see that in both the capacity and speed the human brain is much more efficient and capable that a modern computer.