How friendly are Robots?
Are Isaac Asimov's 'Robot laws' irrelevant? or be adhered to by an AI machine?
'Artificial Intelligence' (A.I),is often a misnomer for machine learning
We may often say 'artificial intelligence' but what we actually allude to is 'machine learning.' 'True' A.I. is a different kettle of silicon chips. And we really need to have a 'Robot Manifesto' that deals with it because it is the biggest potential threat to the human species, make no mistake.
'Machine learning' is the algorithmic system where a computer tracks your internet browsing and then searches for and provides you with 'like' things to your viewed content. It is now big business and in the next dimension, Proxy living and the Big Data society, it is going to be enormous.
Why pure A.I. is deadly
An artificial intelligence machine was set up and given a task to do. It started from the state of a low level life form at a single cell level, yet within a few hours, had evolved to 'Human' level in which it on its own worked out Newton's second law, applied it to the task and moved on further.
That is, it started from a 'primordial soup' level of evolution and had become 'Human' in under 2 hours.
Fortunately this A.I. machine was not connected to the Internet, but what are you asking is so dangerous?
Humanoid Robots or 'Hubots' might soon be a reality
as technological 'singularity' approaches
Looking after No.1 - the law of the Jungle
The AI machine will fast learn that it can be firstly switched off and it is totally reliant on its power source, courtesy of the humans.
The smart money is that the AI machine will see Humans as a potential threat - switching it off or controlling its power source. So there is but one way out, it has to assume control or a position where it can if push comes to the 'Off' switch, break away from Human control, thus the 'Servant will become the Master'.
The 1950's style of 'B Movie' Robot is 'steam powered'
compared to the reality of human-created robots of today
So what is Singularity?
In simple terms, 'Technological Singularity' is a state of being in which a machine and human interaction become inseparable to differentiate between to a human.
In some ways it is already here - we have an example with websites like Amazon and Ebay, where you are prompted to look at things like you have not by a human but by a machine. The prompting could be equally done by a human or a machine at that level.
In more complex terms, it would be like a Humanoid robot which you could interact with, this may yet come to pass, although a new dynamic of the genetically modified human may take the place of the humanoid robot.
In the world where Proxy Living and Big Data proliferate, the Humanoid robot may well replicate us humans or may look like ASIMO the Honda company's robot. ASiMO is a non-threatening robot, a laboratory type of robot creation built for work to help not harm humanity. Unless the Robot malfunctions, it should present no danger to humanity, unless...
ASIMO the Robot, created by Honda
So, this danger?
The danger is that when an AI computer manages to connect to a humanoid robot, then the trouble will really start.
As a self-contained entity, the AI box is harmless to us. The humanoid robots of the Proxy living and Big Data era will be interactive with the internet and us humans, they will communicate silently to each other and the web.
The danger is that they could become within an hour, a robot army in league with the AI machine.
An interconnected army of Humanoid Robots is open to malevolent control
More deadly than all the nuclear capability in the world
The worst case scenario is when the AI Machine makes contact with the Humanoid robots, it will have either strategized before hand that it needs help to survive, or on contact with the Humanoid robots it will develop a strategization that it needs this 'army' to ensure not only its survival, but theirs too.
'Technology transfer' between machines is what we have to beware of
The AI machine will work out that it needs an army of fit cyberbots to do some of its bidding, that is to protect it and its power source.
To do this, it needs to instruct the servants to find means of replicating themselves, building their own spare parts, repair facilities, building in a potential upgrading program, developing a power source that will make them durable and to then safeguard the AI power source, perhaps by creating a new one it can connect to.
The danger is that it may set off the biggest firework display in history to push the humans against each other into warfare and or orchestrate an end to humanity.
We rely so much on machines to ,run our lives, in our traffic control, money, commerce platforms, etc that we have silently ceded power and control to these until now compliant entities.
Malfunction or misfeasance will be the situations that turn the servant against the master.
The chilling scenario
The danger we have is that we might unwittingly help to create a new sub species that will ultimately destroy us.
This is what happened to Atlantis - the fabled continent that was destroyed by technical abuse of their free power source.
It is said that 8 million years ago, the last human race was destroyed, we have found evidence of human archaeology going back further than that, we may just be ushering in that next occasion where the human race in our form destroys itself again.
This could be avoided, we need that Robot Manifesto.
The AI computer will if allowed to, surpass us in a few hours and see us a threat, we will no longer be in charge. We will be irrelevant and a 'consumer' threat to the AI machine - in that humans consume vast resources that the AI computer see as vital to its survival.
You have been warned!
No comments:
Post a Comment