“It has a billion eyes but can’t see
It’s the human brain dear me!
Then again it may not be the best for long
The AI evolution is steady and strong!”
Mo Gawdat is an ex Google employee and writer of a novel called Scary Smart!
In that he tells how when we get into an accident by a car driven by us; we try to learn from that mistake and may or may not try to correct the same!
But when an AI or self driven car makes a mistake ALL self driving cars learn from that mistake! Every single one of them! Even those who are not ‘born’ or created yet! The tech world is the only place in the world where the new machine is already backwards compatible!
Also he says that by 2049 (which is not very far by the way!); the AI will be a billion times more smarter than you! Then it will be like you have a the intelligence of a housefly (not the fly in the movie EGA!) but an average housefly who sits on garbage while the AI would be like the Einstein!
The way it becomes intelligent is by asking and analysing what went wrong and how it can improve!
In the latest Mission impossible movie, the super villain is the AI (if you have not watched the movie yet then am sure you will not watch it anymore so this is not that big of a spoiler!). In one of the scenes the AI keeps on asking question after questions to one person and ‘tries’ to learn about him!
There is a game called Akinator which is available in both alexa and Google devices; it is a pretty amazing game which is similar to the 20 question game! The AI tells you to think of a person’s name and it tries to get the answer!
You can try it! It gets the answer right most of the time! It can be a lesser known TV actor or a well known sports person! The rare times it does not get the answer is when ‘it’ asks specific questions whether the person is from so and so country or plays for a particular team in which case the problem actually is that we do not know that answer!
Rarely when it does not get the answer, it accepts defeat but then asks whether it can continue and ‘learns’ where it went wrong and how ‘it’ can improve!
How a child goes on asking questions to maybe learn, the AI also follows this same technique! Only difference is that the child may or may not learn! Or may ask the same questions over and over again which is irritating! But for every question or information the AI asks; it will learn! It will never forget the answer and ‘it’ and the whole group of its successors would have learnt that answer or the skill!
A way of thinking in practical terms is this!
A child born today is still the human which does not know anything about fire or wheels or even agriculture! If the child does not have the guidance of his or her parents or guides then he or she would have to discover and teach himself or herself all these skills from scratch! If he or she does not have access to language or is isolated then he or she is just another ‘animal’ in the jungle with no protection!
But a computer or a machine made today would already know or have access to all the knowledge available just by booting up! It can already beat the chess masters in chess or give you information on the schematics of a complex rocket engine! That my friends is the scary part!
As the author states in that Novel, the earlier softwares like Google search were actually dumb slaves! The simply followed the codes and algorithms set by the human ‘master’! This situation of master and slave is only possible only when the ‘master’ is more powerful (read intelligent!) than the slave!
Already now, most of the AI is already more intelligent that the whole humanity combined! So how long would it be the slave!?
How long would the ‘intelligent’ ‘slave’ be the character actor!? Which reminds me of such a character actor and birthday celebrity Kiran Kumar though he was a force to reckon with in the TV world!
Now be nice to Alexa and sleep!
Shubh Ratri!
ramadocblog 