There are three basic constituents to the world we know and live in: animal, vegetable and mineral. Generally, we know how it all works together as a sustainable, evolutionary entity – ignoring, for this discussion, the man-made effects upon this planet’s ecosystems. All life, as we know it, goes through a growth stage, reaches some form of maturity, ages over time, and eventually dies, one way or another. That applies to all naturally occurring life forms on this planet and, arguably, on other suitable planets in this universe.
Clearly, robots do not die, in the above sense. Moreover, no robot goes through similar stages as for all animal life on this planet. And no robot has ever spontaneously evolved in nature. But they do cease to function when the power supply is removed, depleted, stopped or interrupted.
So, how can we square that with Professor Brooks’s claim that he is a robot? Without further input from Professor Brooks, I can only speculate…
Well, as already mentioned in my prior posting, the professor and I regard the human body as a biological (or animal) machine. And, we are not alone: the 16th century philosopher, René Descartes, also suggested “the human body could be regarded as a machine”. For sure, there is a crude analogy between the multitude of wires, struts, servos, joints, appendages and so on that make up a humanoid robot and the various types of bones, muscles, sinews, etc that help us all get along in daily life. I could suggest, though, the same analogy could be applied to the chimpanzee or gorilla. Moreover, man-made machines don’t require any of the animal relief we all need, more or less daily e.g. sleep, food, exercise, sex and so on. Hence, although humans and robots are similar by definition, the machine analogy is thin at best and should not be taken seriously as an argument for “I am a robot.”
On the other hand, it’s clear that at birth we have animal instincts and as we grow, we learn to become the person we each are through socialization. In addition, according to Noam Chomsky’s theory of Universal Grammar, we are innately hard-wired to learn and use language; although, some social scientists still disagree. So, perhaps those human attributes can be regarded as a form of natural programming that develops from inception?
In contrast, when fully constructed, a robot requires some form of introduced programming, either as firmware or software before it can do anything. So, in my opinion, the programming similarities are indicative only of the ability for humans and robots to learn; and they are not sufficient to persuade me to agree that “I am a robot” also.
What’s left? Well, there are those who sincerely claim that robotics promises a new kind of life, especially when AI reaches a human-like maturity. If a robot does ever reach the sophistication of fictional entities we can all recall (think Terminator, Ash, Bishop et al), then the claim of new life must be considered and analyzed. The discussion then hinges first on what is meant by ‘life’, ‘living being’ or something similar; and in doing so, the term ‘robot’ for such an entity then becomes problematic.
A comprehensive discussion on the topic of ‘life’ itself is found here, with a list of life forms. There you will find seventeen different categories of life subsumed under Artificial and Engineered, only three of which are directly relevant to robotics: Android (Robot), Cyborg and Robot. None of those, however, provide a clear definition of the term ‘life’ in relation to robotics. Dictionaries and encyclopedias, of course, are useful but not categorical: some definitions do change, over time. So, is any one definition for ‘life’ as good as another? If there is no consensus about a single suitable definition, whose definition do you choose?
To cut through to something more helpful in this topic, I’d suggest considering this definition from one of the world’s greatest thinkers: “One can define living beings as complex systems of limited size that are stable and that reproduce themselves.” You’ll find that claim in the last chapter of The Grand Design by Stephen Hawking (Bantam Books, ISBN:978055381929, p. 224).
That definition covers many forms of life, including humanity, but not robots: no robot reproduces itself. Obviously, no robot procreates as humans do. Moreover, recall that professor Rosalind Picard stated that scientists cannot yet provide four human elements for robots: (1) feelings (or morals); (2) conscious experience; (3) soul, spirit, the ‘juice of life’; and (4) free will – free choice. The last is the most important, in my opinion, because that attribute specifically excludes the term ‘robot’ which, by definition, is an entirely predictable machine that cannot have free will. Hence, man-made machines that do eventually incorporate those four attributes – if that day ever occurs – will require, I’d suggest, a new descriptive word; even the term ‘android’ is insufficient because that is defined as ‘robot’ also.
At present though, when assessing all of the above, and specifically the ‘I am a robot’ statement by Professor Brooks, I can only conclude that he may have been suggesting, obliquely, that machine intelligence in itself truly represents a new form of life, and one that is not yet fully recognized or appreciated. In other words, I think he is saying, essentially, the level of intelligence that exists in your head can certainly exist in a man-made machine, regardless of its structure and mobility. Any other interpretation of the claim fails the scrutiny discussed here.
The implications of that conclusion are quite profound; and beyond my knowledge and experience to fully comprehend, or even accept at this time. I must, though, concede the existence of the real possibility; therefore, the probability must greater than zero.
Appropriately then and as a concluding piece, my next posting will explore some of the clever – or not so clever – stuff currently being done with machine intelligence.