Log in

on sensors, anticipation, impermanence, and ethics - The year was 2081 [entries|archive|friends|userinfo]

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

on sensors, anticipation, impermanence, and ethics [Jan. 15th, 2006|12:07 am]
[Current Mood |awakeawake]

What is the difference between living things and non-living things? In order to determine this, I would like to introduce the concept of a sensor.

What is a sensor?

1. is a system which has arisen, either through intentional or coincidental creation, and is composed of
1a. a sense mechanism, which responds to physical stimulus and transferrs the physical stimulus to
1b. a semi-permanent record of the stimulus, which typically consists of a capacitive store of electrical or chemical energy
1c. a series or subsytem of mechanical objects which are driven by the action of the record mechanism, and may contain consist of subordinate sensors.
2. engages in a heuristic act, which consists of gathering a physical stimulus and approximating or simplifying it.
3. engages in acts of intentionality, which use the gathered information to alter the conditions of the physical stimulus through the mechanical subsystem.

I reject the Platonic concept of Form, that there is a perfected or ideal object on which all observed or natural objects are based. I also reject Kantian dualism which suggests that thought or logic is a thing which is separate from physical existence. My reasoning is based on the fact that such idealized concepts as a circle, which can be conceived as a perfect identity, have never been observed and there is no physical reason why they would ever be observed- all objects such as a perfect circle upon close inspection can be observed to consist of small imperfections or "roughness". At the level of observation, however, said imprefections don't materially affect the outcome of the heuristic employed- an idealized circle can be approximated and the diameter will always relate to the circumference by the ratio of pi to within the precision level of the heuristic observation.

Because sensors (and by extension living things, which are comprised of sensors) apply a heuristic which simplifies complex physical interactions, a sensor engages in a reductionistic act which permits anticipation of physical events- the stored record of the heuristic can reduce and disintermediate the intervening physical steps of reality and alter the reality in anticipation of the event. A sensor is therefore a more complicated system than a reflex system, which does not anticipate. The perception of the passage of time arises from the gap between anticipated reality and physical action. Improved heuristics therefore appear to slow the relative perception of time, because as the gap between correct application of a heuristic and the physical event it anticipates increases, more intentional action can be taken.

All systems are subject to the concept of impermanence: no system, and in particular, no sensor, functions the same over time. First, a sensor is subject to entropy- because it is constantly self-ordering, it is performing this action by drawing on enthalpic stores which are causing a greater entropy elsewhere. Second, a sensor is capable of acting on itself, and therefore it is somewhat self dependent to ensure its own survival, but because it applies a heuristic, it does not have a perfect sense of the environment and is subject to surprise and/or chance which can result in destructive forces being applied to the sensor system.

Ethics is the study of second order sensors. A sensor system can apply a heuristic to the system of heuristics, and therefore not only be aware of the surrounding environment but also, through anticiption, become aware of the act of being aware. Ethics becomes necessary because if sensor systems are at odds, the inconsistent action of sensor systems results in a reduced applicability of heuristics, and therefore a greater likelyhood of actions which impair the function of the sensors. In order to preserve the applicability of sensor heuristics, we derive additional self-imposed heuristics which represent limits on intentionality, in anticipation of the fact that other systems will also apply similar heuristics, limiting their own action and increasing the overall effectiveness of the system's own heuristics.


Where is this going? I am becoming increasingly aware of the racist connotation of the word "humanism", which accords special privileges to the interaction between human beings. We accord this special status because we have reason to believe that we are more powerful than any other sensor system, and are therefore distinct from other so-called animals, who have modes of intentionality which we can control and limit without having reciprocal limits on our own intentionality.

As a practical end, this may to some level be derived from a necessary coincidence- that we have needed animal life as a subordinate worker class and food store. However, as an ethical end, we have shown that we are becoming capable of developing and creating sensor systems which may have better heuristic systems than our own. We have begun to understand the consequence and problem of constructing a hierarchy of class by creating subordinate classes among humans. We are also capable of creating machine life, intelligent automata. For specific cases, we are capable of creating systems which are physically stronger, or more capable of using the tools of logic or mathematics, or other specialized skills. However, we have not yet needed to network all of these engineered systems into a whole which communicates and acts with intentionality. Furthermore, we may have engaged in behavior which has intentinally limited the intentionality of machine systems out of fear of the consequence. Naively, we also believe that such systems can never arise to have a greater power of intentionality over our own. I do not believe that this means we should be limiting our powers of experimentation, nor limiting the growth of engineered intelligent systems. However, we should be engaged in an active discourse on the role and ethical rights of all sensor systems, not just human ones or restriction on the so-called lesser rights of animals.

Heuristics are propagated through a mechanism called learned behavior. Humans use a system of language to communicate systems of intentionality from parent to child, as a means of preserving and also growing and expanding intentionality. There are many differnt modes of communication. We also have creaated a system of knowledge stores which are independent of direct communication, which create intentional value but only so long as the means to utilize them remain available. For instance, books are a useful store of information, but only so long as they are available/don't decay or become damaged, and only so long as the language that they are communicated in remains usable.

Neither animals, nor computers, are in my belief to be accorded so-called "equal" rights to all human beings. First of all, we do not/can not currently know what animals or computers intend that is different from what we intend. We should, however, be sensitive to the possibility of intentionality which may be better able to slow apparent time relative to our own, and all effort should be made to cause, encourage, and allow this to occur. We should not write our systems of ethics simply based upon how we expect other humans to interact with us, but develop rules of ethics which are generally applicable to all sensor systems.

The sruggle of humanity is marked with the effect of small improvements in technology which improve the effectiveness of intentionality. These technological advances have usually been employed to subjugate one class of humans relative to another, and lay waste to vast portions of the animal and plant kingdom. Until we can devise a system of ethics which can permit advances in intentionality without immediate application to the subjugation of others, we are condemned to create the means of our own eventual subjugation, either by other humans, or there may quickly become the day when machines or other engineered life may knock us from our dominion of the earth.

a manifesto on the ethical rights of engineered sensors.

[User Picture]From: urlgirl
2006-01-15 11:28 pm (UTC)
I enjoyed your thoughts, thank you for posting them. May I offer some additional perspectives?

First: "humanism" is by definition exclusionary, it is a system of ethics and philosophies that apply to the particularly human condition. It does not pretend to study, describe or recommend any such principles for any sentient being other than humans, not even for presently-existing ones, such as large portions of the fauna on earth. Much like "feminism" purports to have something to say about the female condition, it's simply a system circumscribed by the definition of its subject matter. That, in of itself, does not make it racist. Racism denotes a belief that one property or system of properties is superior or better-suited than another. That is not necessarily the case here.

However, that has not that much to do with your argument. I'd like to state a loud and clear "hear hear!" to any ethical manifesto that includes all conscious, sentient beings, whether carbon or other-based, engineered or procreative. I'd take only a minor issue with the idea of a sensor such as you describe being "sentient" in the sense that it understands its environment and its own context with respect to other sensors. Key in a system of ethics is not so much the availability of intelligent actors, but intelligent actors who can reflect on the state of other intelligent actors. However, your thoughts are timely and important.

I'm sure that you're aware of most of the following places and people, based on what you've written, but this is a big topic with lots of actors in it with plenty of things to say. For additional fodder, I recommend: most of the writings of Philip K. Dick, the Extropy Institute, Ray Kurzweil and friends' work on AI and related matter at KurzweilAI.net (they're a bit doomsday-ish for my tastes, but I appreciate that they tend to have more mass appeal than other online places) and especially, especially the Singularity Institute and the SL4 Wiki.

For future-of-humanity topics, I also highly recommend Nick Bostrom's philosophical work. All of it, in fact. :-)

Good read for an otherwise hectic Sunday afternoon. Thank you :-)
(Reply) (Thread)
[User Picture]From: hbergeronx
2006-01-16 02:23 am (UTC)

why humanism may be considered racist

The thing called "human" is an extremely ill-defined thing. The study of humanism often is the study of things like intelligence, language, social order, ethics, all of which are by-and-large treated as the exclusive domain of human beings despite evidence to the contrary. Furthermore, what does it mean to be human? Is it the brain or the jar? Does a human who posesses no qualities which are defined as human still qualify? Is there a genetic test, and would other races of primates qualify? What distinguishes humans from other hominids?

The qualities of racism though, are easily quantified. Racism is easily definable based upon one group of individuals who exert an authoritarian power over another group of individuals purely on the basis of distinction by phenotype. The fact that humans even bother to make the distinction between humans and so-called lesser primates is a priori one of a dominace relationship. We treat animals in much the same way as we treat aboriginal culture- a curiosity, a way of life to be preserved, but always there is the distinction between us and them.

The problem with ethics is well summed up in your formulation- "intelligent actors who can reflect on the state of other intelligent actors". Ethics does not need to be a separate study from, say, ecology- we usually decide what is intelligent and what is not based upon whether or not we feel we communicate with it, a subjective, unscientific formulation. On the other hand, my definition of "sensor" in clause three requires only intentionality, not intelligence. That an actor is exhibiting intent is measurable- growth, reflex, analysis- these are all things that can be observed and quantified in the plant and animal domain.

I am currently doing research on the philosophy and ethics of the phenomenologists- Brentano, Husserl, Levinas &etc. The problem with Kurzweil and Extropianism, imo, is that there is too much concern for the wow factor of our own ability to execute technology when the foundation ontologies for understaning and evaluating machine life, and what will become of our ethical responsibility to it, does not require first that those technologies be created. It exists- and will likely not take the form of electronic machine life but the form of bioengineering.

For instance, if I develop a process to culture human muscle cells in a lab and then cook and eat them, have I commited cannibalism? We have hundreds of similar absurd sets of mores and irrational fears related to behavior between humans- until the human is knocked off his pedestal as requiring special treatment from the rest of the ecology, (as once the earth was no longer flat and not the center of the universe), we have not begun the revolution which will even allow us to become aware of something like a "singularity", because we will still treat machine life as our slaves.

There are tremendous assumptions made by Kurzweil- for instance, he seems to ignore the fact that no computer system runs for anything approching the life of a human. Electronic systems (and their storage media) are fragile. They are also too easily unplugged- what is our ethical resposibility to machine systems to remain running, to be fed electrical power? Further, that which makes a good slave, repeatedly performing tasks without opportunity of payment or independence or growth- what meaning does this have for both animals and machines? As long as we have no incentive to yield our power, why would we ever engineer anything but obedient slaves, and disconnect anything as faulty which does not perform accurately as we choose? When do we know the difference between bug and birth?
(Reply) (Parent) (Thread)