[Fis] ON INFORMATION ETHICS - Rational Machines

[Fis] ON INFORMATION ETHICS - Rational Machines

From: <[email protected]>
Date: Tue 07 Mar 2006 - 14:03:28 CET
10th FIS Discussion Session: ON INFORMATION ETHICS Answer from the Viewpoint of Rational Machines The questions to open the discussion were: Q1. What is the philosophic and ethical challenge of the physical, chemical, biological... sciences today with regard to human freedom and the 'laws' (?) that should regulate our actions? What is the meaning of 'natural law' today and can we take it as an analogy (?) in the field of morality? are there other alternatives? Q2. What is the philosophic and ethical challenge of modern information technology with regard to human freedom? How far do we conceive the 'cyberspace' or, more generally speaking, the potential digitization of all phenomena (human and non human) as a (the?) condition for understanding them, and what does this mean with regard to human behavior? More specifically: do we conceive ourselves eventually as information processing devices?, and if yes, what follows with regard to artificial digital devices that we are creating now and in the future? Q3. What does it mean for human beings to be able to behave in a global digital world in which time and space seem to disappear? What are the consequences of our doing for the whole planet? How can we create rules of action that are accepted by all human beings in order to achieve a global sustainable (physical and cultural and economic...) development without deleting all differences that makes human life worth living? In case we would be able to construct an Artificial Person, we would have to address these points, too. So, as we are � conceptually - able to simulate life as a mathematical process, we are confronted with the consequences of the extent of ethical attitudes within the consciousness of this living organism. An ethical rule is in an autoregulated system a rule of behaviour. (It does this because it {usually, normally, always, �} believes this to be the right thing to do.) We deduct an ethical system from the actions of an agent. Therefore, if we lay down rules of behaviour for an artificial intelligence, we lay down its ethics. Asimov has looked quite deeply into this philosophical question (in his theory of robots, as expressed in many of his works.) If the living organism is an abstract cell, we could think its ethic to be directed towards survival, growth and reproduction. �Recreate and multiply� is a quite strong, basic, ethical rule. In a more complex environment, optimality of reproductive success can be achieved by further rules, which can be subservient or concurrent (in different hierarchical relations). (e.g. in order to avoid retaliatory extermination, do not kill unless in self-defence) This discussion gives us room to design the ethics of an artificial intelligence. Q1. What is the philosophic and ethical challenge of the physical, chemical, biological... sciences today with regard to human freedom and the 'laws' (?) that should regulate our actions? What is the meaning of 'natural law' today and can we take it as an analogy (?) in the field of morality? are there other alternatives? A1: In this, or in one of the very next, generations we shall be able to fuse the �rational� concepts with the �emotional� concepts (mechanics and biology). We shall find a gearbox which translates rational relations into congruence with relations discovered on biologic observations. Then, the two sets of connotations will have to compromise. Either, mechanics will also be more divine-transcendent, consciousness-aware, or biology and psychology will be more rational-economic, statistical-optimising, predictable. If an agreement is reached that biologic organisms (including humans) act and react along rational rules, this has the consequence that the shamanic class of services suffers a crisis of legitimacy. If any thinking person can deduct the rules of good behaviour, there is no need for specially anointed transmitters of the divine (irrational, non-understandable, transcendent) Will. Q2. What is the philosophic and ethical challenge of modern information technology with regard to human freedom? How far do we conceive the 'cyberspace' or, more generally speaking, the potential digitization of all phenomena (human and non human) as a (the?) condition for understanding them, and what does this mean with regard to human behavior? More specifically: do we conceive ourselves eventually as information processing devices?, and if yes, what follows with regard to artificial digital devices that we are creating now and in the future? A2: The CNS (central nervous system) is an information processing device. The question is, whether one thinks himself outside his CNS. Current thinking equates �consciousness� with electrical processes (of the alpha-kind, mostly) that take place on the brain while the subject is awake. The veterinary part of the human mind is currently conceptualised as the collection of the biochemical processes that interact with the electrical processes. By and large, one can simplify �emotions� and �thoughts� into biochemical vs. electrical processes. These co-regulate the brain. Freedom is, statistically, unpredictability. The extent of predictability is given by Nature (as long as we believe the natural numbers to be a gift of Nature) through some numeric interweave-patterns that explain the co-regulation of a quasi-fluid matter emitting electrical bursts. The case of a human making a decision based on his free will is a special case of the predictability of the properties of the next state out of the properties of the present state. How fixed is the future, generally? The answer to this lies in the structure of our basic concepts. (If one grows up in a very strict environment, his opinion of the world will be that it is a non-chaotic system.) We learn our basic concepts - on the rational level � as we go to elementary school. If the (system of rational concepts about the) world has been presented to us as a place where nothing happens by itself, we shall have a static idea about the world. (see: Newton and the non-existence of a perpetuum mobile, e.g. We ourselves experience ourselves as non-static and own-triggered and a perpetuum mobile � maybe only outside school.) Trying to find a compromise between the Newtonian idea of a natural, predictable state (idle, etc.) and biology, which is anything but idle, and in some fashions less, in some fashions more predictable than the non-living nature, one finds that predictability is the interplay between what is now and what will be. Counting now, how many diverse kinds of �now� there are and how many diverse kinds of �will be� there are, one finds that the interdependences are more-dimensional and quite complex. The main challenge for information theory is to create an artificial organism that has moods (urges, needs) that can be satisfied by more than one possible action. The good news is that the problem can be solved. Q3. What does it mean for human beings to be able to behave in a global digital world in which time and space seem to disappear? What are the consequences of our doing for the whole planet? How can we create rules of action that are accepted by all human beings in order to achieve a global sustainable (physical and cultural and economic...) development without deleting all differences that makes human life worth living? A3: Human behaviour is deeply irrational, instinct driven and opportunistic-maximising. The irrationality refers to stated aims and stated expectations. It is highly rational with respect to the production of hormones within the person. The instincts that govern human behaviour are, as expected, just as ethology describes (rivalry, territoriality, nurturing, etc.). The matter (hormone) human systems maximise is �congruence� (maybe, neuropharmacologists call it endorphins.) The congruence between �expected� and �experienced� values within the CNS is what drives the human animal. (In German, the terms Sollwert and Istwert render the idea very tellingly.) As the lasting maximal congruence between Sollwert and Istwert cannot be achieved (due to some number theoretical voodoos), there can be no winning strategy (specifically no single winning strategy). If there would exist a good (or slightly better) strategy how to live (procreate and multiply), we would see it everywhere, winning. So, it is useless to look for The General Solution. It is puzzling whether this or the next generation can build an elite among those who can communicate (via the internet) and whether this elite will achieve a critical mass to try to impose (sell, educate) its ideas on the whole of the population. One should not overemphasise the internet. Now we have recreated the situation in Athen where the well-educated and open-minded burghers spent time chewing philosophical questions like we do. They were within hearing distance to each other and could interact on the spot. This is the state we have regained. We sit again on the agora and hear each other an! d can reply immediately. Let us hope we are as well-motivated and investigative-minded as they were.


_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Tue Mar 7 14:00:37 2006


This archive was generated by hypermail 2.1.8 on Tue 07 Mar 2006 - 14:00:38 CET