Wednesday, July 13, 2005
Yoo hoo Alaska!!
I had been to Alaska a couple of weeks back ...was a wonderful trip...will write down a trip report or something in the coming days.
Monday, February 14, 2005
Long time no see
Coming back to my blog site after a long time...
Need to spend some time to pen down my thoughts...
Need to spend some time to pen down my thoughts...
Wednesday, January 19, 2005
The Machine Evolution
Hi,
I saw this movie ...I, the Robot....yesterday....and there was this line in the movie where the creator of the Robots, one Mr. Lanning, says that random pieces of code could, in theory, band together to create more intelligent software....in a sense...machine evolution...
And today...I was reading an article on software programs...where it said that each layer in a software model is an abstraction...
For e.g., in Unix, when you type the command "rm *", we expect all the files to be removed....
So the OS provides an abstraction where we expect that a command entered would be executed accurately....
Now the OS converts that command into machine language or assembly code, which provides another layer of abstraction...where the OS expects that certain segments of machine code would do the job accurately...
Now the machine code is translated into stuff that switches on and off logic gates....so the NAND and NOR gates provide an abstraction that the machine code expects will work properly...
And finally, what is a bit, a 0 or a 1??... Logically, yes....But technically, its just voltages...
If the voltage crosses a particular threshold, say 5 Volts, its a 1 or its a 0....
So you see, the entire software model is nothing but abstractions at each level....
The abstractions can only promise certain things that it will do....it tries to be generic to a very large extent...tries to cover a lot of angles....and it does that....
But....its still possible...that certain sequence of actions could prove to be too much for the generic case at any of the levels....and that is when bugs (either software bugs or hardware bugs or both) appear...
What I try to prove from all this is that its possible to have random behavior from machines...
Now, let me jump to the living world...lets look at the way humans or for that matter other species evolved....
It is widely believed that the animal species and others evolved from single cell bacteria...
Now some of them had certain genetic sequences altered randomly...and such cells banded together and finally evolved into something else over thousands or millions of years...
Comparing these two scenarios, isnt it possible that since we have so many millions of software programs running worldwide, that somewhere... sometime, certain random events could lead to random sequence of codes banding together to form something intelligent??
If we look forward... say a thousand years....isnt it possible that these very software that we have today would in all probability have a "mind" of their own???
That leads me to another thought....
Could it be possible that we are the final evolved species of certain genetic codes that went wrong thousands or millions of years ago???
If so, was there some one writing genetic programs millions of years ago??? I know all this sounds absurd....but what the heck...this is how the world has made progress....
Ok..its time to sleep now .... lots of things to ruminate on while sleeping....
Good night every one....
I saw this movie ...I, the Robot....yesterday....and there was this line in the movie where the creator of the Robots, one Mr. Lanning, says that random pieces of code could, in theory, band together to create more intelligent software....in a sense...machine evolution...
And today...I was reading an article on software programs...where it said that each layer in a software model is an abstraction...
For e.g., in Unix, when you type the command "rm *", we expect all the files to be removed....
So the OS provides an abstraction where we expect that a command entered would be executed accurately....
Now the OS converts that command into machine language or assembly code, which provides another layer of abstraction...where the OS expects that certain segments of machine code would do the job accurately...
Now the machine code is translated into stuff that switches on and off logic gates....so the NAND and NOR gates provide an abstraction that the machine code expects will work properly...
And finally, what is a bit, a 0 or a 1??... Logically, yes....But technically, its just voltages...
If the voltage crosses a particular threshold, say 5 Volts, its a 1 or its a 0....
So you see, the entire software model is nothing but abstractions at each level....
The abstractions can only promise certain things that it will do....it tries to be generic to a very large extent...tries to cover a lot of angles....and it does that....
But....its still possible...that certain sequence of actions could prove to be too much for the generic case at any of the levels....and that is when bugs (either software bugs or hardware bugs or both) appear...
What I try to prove from all this is that its possible to have random behavior from machines...
Now, let me jump to the living world...lets look at the way humans or for that matter other species evolved....
It is widely believed that the animal species and others evolved from single cell bacteria...
Now some of them had certain genetic sequences altered randomly...and such cells banded together and finally evolved into something else over thousands or millions of years...
Comparing these two scenarios, isnt it possible that since we have so many millions of software programs running worldwide, that somewhere... sometime, certain random events could lead to random sequence of codes banding together to form something intelligent??
If we look forward... say a thousand years....isnt it possible that these very software that we have today would in all probability have a "mind" of their own???
That leads me to another thought....
Could it be possible that we are the final evolved species of certain genetic codes that went wrong thousands or millions of years ago???
If so, was there some one writing genetic programs millions of years ago??? I know all this sounds absurd....but what the heck...this is how the world has made progress....
Ok..its time to sleep now .... lots of things to ruminate on while sleeping....
Good night every one....
