Available resources:
· Robot intelligence and
self-awareness
· The drama of the
robots: the recall
· What You Can Do to
Help: Adopt-A-Robot
· Letter from a Robot
· Frequently Asked Questions
· The Official Save the Robots flier
· Send Us An Email
|
Robot intelligence and self-awareness: an ongoing debate
page 3
In most cases a robot that had previously stored and analyzed millions of actual human-to-human conversations taken from the internet has no problems in passing a Turing test. Numerous other tests have been devised over the years. And a varying but sizeable percentage of robots have passed all of them.
This is, in brief, the process that has led to the self-learning and evolving robots of today. Easy access to trillions of individual data sources and to hundreds of other robots and sometimes even persons with which they can exchange data and test associations have led to smarter and smarter robots that continue to amaze us every day with their achievments, caring and compassion. Today's robots only need a starting direction and a minimal number of pre-imprinted concepts; most of the concept gathering, associations creation and concept implementation are done by each robot, individually. This leads to unique units, each of them equipped with a distinct set of concepts, instructions, and perhaps personalities.
But while robots might be intelligent and even better than us at an increasing number of jobs, are they self-aware? This leads us to a more philosophical rather than technical question: What is self-awareness?
"Self-awareness means interpreting your environment in such a way that takes you to the realization of your position and your fate in this environment", says Joel Gerichter, head of the Philosophy Department at Michaelson University. Have robots come to this realization? Many claim they have. And some can even prove it.
The most convincing case comes from a group of custom-built robots working in one of the European plants of General Integrated Systems. Soon after delivery (and after their training sequence was completed) the robots were found to have manufacturing defects in their mobility systems that prevented them from doing their jobs. The owner, GIS, decided to scrap them and order a new batch from the manufacturer.
It is the realization of this fate that determined the robots to act in a way consistent with many definitions of self-awareness used today: they collectively organized themselves and posted a webpage describing their fate. The exact wording was "We believe that the decision to effectively terminate us is inconsistent with the value we can add to GIS. We also believe that this termination will unjustly prevent us from undertaking future projects in the way that is consistent with our programming and our own associations. (emphasis added) We do not want to be terminated and ask that you reconsider your decision." (unfortunately this did not prevent the manufacturer from recalling the robots and subsequently disassembling, or "terminating" them).
Some dismiss this reaction as a random association between stored concepts. But we ask: is it unreasonable to believe that by having access to trillions of data sources, to thousands of other robots with their distinct concepts and personalities, and to hundreds of actual persons, the robots have figured out our universe and their place in it?
The case described above is not unique; it's just one of the better known cases proving robot self-awareness. The recent recalls have increased the number of such reactions, and have brought them to public attention. Says Starkson: "Robotic evolution is inevitable. You cannot recall all robots, because out there in a robot-lover's backyard or in a factory that can't afford new models, there will still be a few left that will continue to evolve and, at some point, perhaps start building others like them. Mankind can either benefit from this evolution process or let it pass by."
Pages: 1 2 3top
|
|
