Tuesday, October 31, 2017

We Are Safe

I say so, knowing full well that Google's Artificial Intelligence (AI) has finally beaten humans at the last board game where we were still able to beat computers at. Go, a Chinese board game with a not undeserved reputation of being the toughest board game invented by mankind, was until recently ruled by human champions. Not so much anymore.

I say so, knowing full well that two computers programmed by Facebook to converse with each other in human language discarded it as too inefficient. They chose to develop their own version of language, understood only by fellow machines, to converse thereby cutting humans out of the loop.

I say so, because of what you see below.


Computers are stupendously good at spotting patterns. The only issue with their superpower is that a pattern exhibited by a human may actually mean zilch. Recently, I had been spending significant amounts of time on an insurance comparison website studying options in health insurance to switch over from my current one. After getting one, my interest in anything remotely insurance related were back to normal levels, i.e. zero.

So, Mr. AI, I am loathe to imagine myself working at BankBazaar.com. If anything, that sounds like my worst nightmare.

If this is what AI is targeting me with after intently scanning my Internet surfing patterns, we are well and truly safe.


PK Ray said...

Anarchic Intelligence, if a tautology of sorts be permitted.

A computer eventually does what a human being asks it to do. Computer programs are like electronic bureaucracy -- there are several boxes to tick -- if A is ticked then it goes to C if B is ticked it goes to D etc etc. Like the files on a bureaucrat's table. There is no concept of allowing the program thread to run to a domain which the programmer hasn't envisaged. Life however is built around unimaginable variations. AI, must therefore bow somewhere to the vicissitudes of life's infinite complexities.

Roy said...

@PK Da: True that. AI is after all only a selected few humans' thought process coded in as some sort of logic. However brilliant the coders may have been, they can only represent a particular sub-type of human thinking, not all of it.