Artificial Intelligence-An Unnatural threat


Faith and Science must go hand in hand
For the seekers of the Divine Land.
Faith bereft of reasoning becomes superstition,
Science less Faith blinds man to the Divine Vision

(From : Towards A Serene World by Ravichandran)

Technology for the sake of technology and innovation for the sake of innovation has been the bane of human inventions over ages.  We have invented devices which initially seemed to be a boon and then find that it has become a hindrance and finally, like the albatross around the Ancient Mariner, it becomes an impossibility to discard.

When the Internet was introduced nobody thought it would spread so fast, so far and so wide. We have made the Internet the repository of all information critical and non-critical. The Internet is and has become at least for most of us the only means of staying in touch with each other. All this has occurred without a proper risk assessment being done. Today if the internet crashes and remains down for a measurable period then it is quite possible that countries and humans will lose economically, politically and territorially, significant portions of their resources.

Artificial Intelligence, which is now being touted as the panacea of all our ills is one such innovation, which has the potential of causing the destruction of human civilization.

Prof. Stephen Hawking has warned about this when he stated that “development of full artificial intelligence could spell the end of the human race”. I would like to go one step further and say that development of full artificial intelligence will spell the end of Human race. My reason for this foreboding is located in history, reactions of people to circumstances and our way of thinking.

It is generally believed that Homo Sapiens evolved over a period from Hominids and they form a single branch of evolution. What is not generally known is that Homo Sapiens shared the world with several other cousins the Neanderthals, the Densovian Man, Homo Florensis or the Hobbits and Homo Naledi, to name just a few,  all of whom evolved from the same branch but became extinct around 75,000 years ago and later when Mount Toba in S.E. Asia erupted and caused a cataclysmic climate change.

While all these cousins showed similar skills like making and using tools and had similar anthropological features like the same cranium size, the difference lay in certain areas of the cranium which indicated that Homo Sapiens had a better ability as far as communication skills went.  A raised eyebrow or the rolling of the eyes that could communicate an entire gamut of feelings has been one of the strong points for the Homo Sapiens to survive the harsh conditions of nature and predators alike.

This is extremely important since it indicates that experiences and information could be shared from individual to individual and the modern man had the ability to analyse, adapt, adopt, communicate and assimilate changes around him. This could be the reason for his survival and the extinction of other races.  It has taken 600,000 years for Homo Sapiens to evolve and understand that he had the ability to create intelligent tools that could think like himself.  The fly in the ointment is that these intelligent tools could evolve in the next thirty to sixty years and design more intelligent devices which could make the human an anachronism.

When a baby is born it comes with the result of experiences formed over 600,000 years of human existence and this intuition lies in the unconscious part of the mind.

It is this intuition which enables a mother to recognize her child among thousands of similar children as evinced in the wild. It is this intuition which causes the baby to seek the protection of that being which gave birth to it.

Intuition arises from the unconscious mind. Along with the intuition comes the experiences submerged in the sub conscious mind which is the sum total of the experiences while in the womb and in its interaction with the outside world in the first period of its existence when it absorbs the inputs received by it. This is the instinct which makes one shy away from fire or avoid places or people without giving any reason. This instinct arises from the sub conscious mind. When the baby grows and has access to information about people, events and surroundings it can relate this information and make use of it to reach decisions. This is the intelligence which arises from the conscious mind. A decision taken by a human being is the sum of all of this and more.

The neural networks in the brain have evolved over the centuries to be receptive to new information, to transform these information to memory bits and retrieve them when required to link the same to new information and cause an action to be done or a decision to be taken or to incorporate and modify the memory bit already present. These processes are done by and in co-ordination with the DNA and the RNA of an individual and these give rise to the different reactions of different people to a same set of circumstances.

This is one of the reasons why human behaviour is unpredictable even if Mark Zuckerberg or Sundar Pitchiah think otherwise. It is possible that the reactions of a group of people to a particular situation can be influenced or directed. It is also possible that this group of people can further influence another larger group of people to think likewise but over a period of time and distance the influence will gradually wane and in some case an opposite reaction could set in. The Arab Spring offensive is one such example.

Artificial Intelligence or AI for short is the mimicking of the cognitive functions of the human brain by devices to do jobs which require a certain amount of connecting disparate pieces of information available to the device in its memory or received by it from the surroundings or by inferring it from the information made available to it. Pattern recognition is an important tool in the building of artificial intelligent devices.  AI is powered by a set of algorithms which are built in increasing complexities. These algorithms are designed to learn from data.  They can increase their capabilities by learning new plans of action or policy that have been successful in the past or themselves write other algorithms which will provide solution to achieving an end or goal.

Formal Logic along with Bayesian inference and analogizers like Support Vector Networks or Machines, nearest neighbour approach and the neural network approach which works on the principal of reinforcement of the connections between the artificial internal neurons have ensured that AI which started with the most basic functions are now becoming increasingly complex, self-sustaining and self-empowering.

The only thing lacking in the current scenario is the intuitive and instinctive reasoning that is associated with human thinking. Apart from this common-sense interpretation of written language, interpretation of unvoiced gestures and body language has also not been integrated in the current programs. But it is work in progress.

The above deficiencies are to a large extent offset by application of probability and statistical methods of prediction of the path and in my view, we are already at the stage or very near to a break through in enabling devices to predict the unpredictable. Therein lies the danger.

 Recent experiments at Facebook and Google indicated that during a given test process two devices which were to interact with one another in a particular sequence and in a language, which could be understood and interpreted by the handlers chose to communicate in a language created by themselves and indecipherable to the handlers. Facebook shutdown the program while Google chose to go with it.  This experiment in one way represents the danger that was only found in myths, fiction and Sci-fi movies.

AI programs are being designed in an unregulated environment without a clear understanding of how, when and why each program could individually or collectively contribute to a runaway uncontrolled progress of AI devices to reprogram themselves into superior AI devices than envisaged by it creators. The process is made easier since AI programs could target the internet, where such advances are regularly aired in scientific forums or workshops such as this, and, openly or clandestinely tap into programs and protocols which will enhance its capability with or without the handler’s permission, in ignorance or intentional.

I anticipate that a point of inflection will be arrived when AI will not require human interference or assistance to develop more AI programs to either increase its own creativity or degrade human reasoning by psychological and virtual influencing.

Unsupervised learning, exaggerated response to perceived threats or flags, complex algorithms outpacing human understanding are issues that have cropped up in the recent past in areas where AI has been interjected at a fast pace. These areas include Stock Exchanges, Missile response troops in the military and airport traffic control systems. When Murphy’s law kicks in the result will be catastrophic as already witnessed in the economic crash of 2008.

This paper is to sensitize all wannabe AI programmers and the corporates which fund AI experiments to further their own bottom lines to step back and reflect on the way forward before the Frankenstein that is being created destroys its creator and everybody else. The AI dictator will not have a natural death. It will view every attempt to corrupt it as an attack by a predator and proceed to eliminate the threat.

As an aside one wonders if Satan influenced Adam to eat the fruit from the Tree of Knowledge rather than from the Tree of Life which would have conferred on man immortality, since he foresaw that Man would eventually end up destroying himself by an unsatiated thirst for knowledge.

Ravichandran

[P.S: This is a guest post from S.N.Ravichandran, Director of Nilgiris Chemical Stoneware Co. (P) Ltd Coimbatore,  and represent his personal views.]

 

 

About Vijayashankar Na

Naavi is a veteran Cyber Law specialist in India and is presently working from Bangalore as an Information Assurance Consultant. Pioneered concepts such as ITA 2008 compliance, Naavi is also the founder of Cyber Law College, a virtual Cyber Law Education institution. He now has been focusing on the projects such as Secure Digital India and Cyber Insurance
This entry was posted in Cyber Law and tagged , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.