Saturday, October 1, 2016

Attachment to Robots and Intellos in Alien Civilizations

In trying to figure out what life would be like in alien civilizations that possessed the technology and capability to travel to other stars, and potentially visit us, one aspect that came up was training. Recall that training is the predecessor to education, and is used here in a global sense as well as in a formal one. We humans start training ourselves as soon as we are born and conscious, as babies are not born with much ability to move, recognize faces or foods or other important items, or actually anything at all except suckling. That is enough to continue life and allow the magnificent neural nets we are born with to start processing information and correlating it. Humans are about at the limit of what a newborn creature can have automatically and still survive, meaning that all we know is from training, first by ourselves and then by others, and then from education. For maximal intelligence, this is probably a mandatory step, and so we would expect aliens on a species that attains high intelligence to be similar. Almost nothing programmed in, but a tremendous potential for our neural nets to correlate, see relationships, categorize and then recognize.

Call it interstellar convergence in brains if you wish, or simply a necessary condition for high, flexible intelligence of the kind needed to create civilizations and technological advances, aliens would be similar. They would not be like insects with instincts programmed in, as the genetic coding of behavior and other abilities is extremely inefficient at generating intellectual capability. This means there would be a long-period of self-training, probably guided, followed by more and more guidance in training.

It also means that any object in the field of view, auditory envelope, or tactile touch-space of a young alien is going to be incorporated into their neural nets. If robots or intellos, which as a reminder, are the less-expensive genetic equivalent of robots, are involved with interacting with young aliens, they are going to be in their attachment list, which are objects that generate positive emotions. If a young alien is fed by a robot, that robot is going to be associated, at a deep neural level, with positive feeling. Feelings are simply brain responses to situations that have enough familiarity to trigger recognition in a neural layer that is associated with positive rewards, such as food. The young alien is going to have feelings for that robot, and perhaps other robots that are similar to it.

If the alien civilization is not using robots for this, but grows intellos, intelligent creatures that perform functions in the society, to do it, then the attachment will be to intellos, notably the ones that are involved with feeding, cleaning, and otherwise caring for the young alien. With a starting neural net having few instincts, care for youngsters is going to be extensive, with overwhelming opportunities for attachment to grow very strong.

The alternative for the alien civilization is to use adult aliens exclusively in the early care of the young aliens, avoiding this attachment problem. This is possible, but will involve a large block of time from alien adults. Would the adults be willing, or could they be made to be willing, to do this, despite the pervasive presence of robots and intellos in their civilization?

There are many aspects of this question. First, after the genetic grand transition, and its subordinate transitions such as the neurology one, aliens would have all the genes needed for high intelligence. But high intelligence would be wasted if the training they received were not sufficient to boost their thinking abilities to the maximum. It has been assumed that alien genetics would be improved, in intelligence as in everything else, to the best possible. Other areas of improvement would be health, athletic ability, stamina, immunity, ability to recover, sensory areas, communication ability, and more. The adults that might be enrolled as trainers would be equally gifted, but would that be sufficient, or would some robotic artificial intelligence be needed at some stage of the early or late training?

Second, lifetimes would be extended, meaning that the fraction of an adult alien’s life that might have to be spent on training young aliens might be smaller because of it. On the other hand, with larger intelligence, perhaps larger brains, the period of training of young aliens might also be longer, increasing this fraction again. There would likely have to be one-on-one attention to allow the attachment scheme in a neural net to function, so that an adult alien would have a large fraction of their daily lives involved in this task, as opposed to a small amount of drop-in attention. Eliminating robots and intellos from training would appear to be a large demand on adult alien time.

The society that adult aliens live in would be extremely interesting, so the alternatives to young alien care and training would likely be more attractive and much less demanding of the dedication of large block of an adult alien’s life. Would it be possible to make this task desirable for adult aliens? If it was undesirable, and aliens were forced into some sort of involuntary servitude, then the training would reflect this, and certainly there would be some spill-over effects. If adult aliens disliked being the training guides and care-givers for young aliens, this would be detectable early on in the lives of the young aliens, and would have an effect. Successive generations might have stronger and stronger negative reactions to this burden and demand relief by the use of robots and intellos.

Third, what would be the effect on an alien civilization if the average adult alien felt a strong attachment to robots and intellos. Note this is not an intellectual or logical conclusion, and they all might well understand that robots and intellos were disposable parts of the infrastructure. Instead, it is much stronger than these conclusions, and represents the same feelings that one alien might develop for another alien, which is also based on the same neural net operations, leading to an attachment. Would this type of attachment move the civilization to treat robots and intellos as somehow the equivalent of alien citizens, or perhaps to have a lesser tier of rights or privileges, or perhaps to be not simply disposable items? What large-scale effect would this have on the alien civilization? Would they find themselves devoting more and more resources to robotic maintenance and intello health? Or is it solvable by other means?

Perhaps one means of dissolving such attachments to robots and intellos, short of dragooning much of the alien population into care and training of young aliens, would be to make sure there were uncountable numbers of clones of both the robots and the intellos, so the young would develop a different type of attachment to the robots and intellos they were involved with. How does a neural net attachment change if there are thousands of identical copies of the robot an alien became attached to? Does this eliminate the attachment?

It appears this topic, while perhaps critical to understanding an alien civilization’s training and therefore preferences, is very complex. Some further breakdown is likely necessary.

No comments:

Post a Comment