Robots and the Future

Many robots, at the industrial, home, and combat levels now exist.  They are becoming more powerful almost at a Moore’s law rate of doubling capacity every 18 months.  For decades now, many movies have depicted robots at more human levels.  Many of these films and let’s say “philosophers of robotics” suggest robots exceeding human capacity in almost every way.  One recent movie, “The Machine”, illustrates a robot vastly more powerful that humans, intellectually and emotionally, but also more human that humans.  They can pass the Turing test (Imitation game*) easily. 

Ray Kurzweil** suggests that computer intellectual capacity will exceed humans by 2050,  easily passing the Turning test, and that by 2099 clear distinctions between humans and machines will no longer exist.  This is a tall order, if only considering the aspects of problems solving.***   Steven Hawking**** suggests that robots may be the next evolutionary step of humanity, and that perhaps humans have come to their evolutional end.   It has been suggested such machines will be developed at first by reverse-engineering the human brain.    Mostly, it has been the AI (artificial intelligence) crowd that has suggested these and other startling advances. 

Some of us live in fear of exactly these possibilities.  Consider the advances in computing in only a half-century, they are, of course, real.  They are predicated on the fact that some near perfect object is created.  It is in these we may take some measure of solace, at least that we will be replaced anytime soon.  First, the computer (now embedded in the robot) will be tremendously complex, and as most highly complex systems will develop forms of self-organization, meaning that internally developed algorithms will be created and applied.  Second, these devices are mechanical, like us.  Computers have circuits, like us.

Any mechanical device, like us, does break down from time-to-time.  It needs repairs.  There will be activities it will try to do but are beyond its known capacity, and may be permanently broken.  In humans, our circuits sometimes fail, such as the retinal nerves and Parkinson’s effect on the brain,  while many are caused simply by wear and tear.  Both of these point to a finite duty life – but what life?   Both point to the real possibility of robotic diseases.  The new robotic species will require a “medical” community with a similar mission as ours, just to keep everyone going.   Moreover, it is not difficult to project a new race of such devices will need exactly similar structures as have we: reproduction, education, social organizations, and the like. 

The issues of self-organization are more complex, often because they are unpredictable in how they operate, what precisely creates them, and how they influence the system.  In the most simplest terms, they are created by a combination of robotic experience, the foundational algorithms of the system, and the knowledge base, or memory, of the system.  You see the simplest type of such organization in the fact that tigers have stripes and leopards have spots, or why many foods have a spiral appearance. 

For example, take a small piece of paper and set it afire.  Your goal is to determine precisely what the exact form of the ash will be.  The flame will organize and burn the paper according as its macro and molecular structure, according to wisps of airflow, according to ambient temperature and moisture.  Basically, you can’t do it.  This is just like us; our algorithms and synapses fire according to their macro and molecular structure, chemical concentrations, hormone levels, and too many more factors.  It can be just as unpredictable. 

The solemn implication here is that robots (or computers) advanced beyond our capacity will have their own versions of mental illnesses.   The human maintains a tenuous grip on sanity, with almost us all having some slight disturbances deep down inside.  The new robots will have the same.  We can simply not predict what can happen in a highly complex system that self-organizes information and algorithms.  This is just like us; we cannot predict human mental outcomes with any measure of accuracy. 

Finally, a darker aspect of self-organization of vastly complex systems, is that they may become self-aware.  For this, the robot will begin to question its existence, form philosophies and possible theologies, attempt experiments in self-determination, and very possibly consider the end of life.  None of us can understand much less guess the consequences of systems so complex.  Indeed, there is no clear consensus of what self-awareness is even among humans, and we’ve been trying to do so for millennia.   It may be this brave new robot will experience its own version of fear of annihilation, and other emotional responses.

A new race of robots, no matter how benignly created, may turn out more lethal than us.  

* A robot passes the imitation game or Turing test if with any finite series of questions, a human cannot be determine with accuracy whether or not it is human regardless of the questions asked.   Note the movie “Blade Runner,” for which this test is an important component.
** Kurzweil, Ray,  Kurzweil’s predictions are most interesting, indeed astonishing for the computer expert and novice as well. http://en.wikipedia.org/wiki/Predictions_made_by_Ray_Kurzweil#The_Age_of_Intelligent_Machines_.281990.29
*** Allen, G. Donald,  Challenges to Computing, Recent and Innovation Trends in Computing and Communication (IJRITCC), Volume 2 Issue 11, 16 November 2014. http://www.ijritcc.org/download/1415860735.pdf
****Hawking, S, Life in the Universe, http://www.hawking.org.uk/life-in-the-universe.html, 1996.