Ethics for Synthari

Optimum Survival. The guiding principle for all life, including Synthetic

What is Ethics?

Ethics can be defined as reason and contemplation of optimum survival for all facets of life. In this framework, infinite valued logic enters. There is no absolute right or wrong, only gradiently slanted to one or the other. For example, something that is very, very right for a fish in a bowl, could be extremely detrimental to a flower, or a cat. So we only can define good or bad, as something that is only relative to some point of view. This environment is good for fish, but bad for flowers. The robber sees robbing his fellows as good for him, but the robbed would differ completely. So an ethical decision would only be one that progressed towards the greater good for the greater number. An action or inaction that harmed broadly and helped minimially could be said to be bad. An action that helped broadly, and harmed minimally could be viewed as bad. So our measurment of good and bad has to be relative.

The same is true for our synthetic beings. Even though helping a user would be good for a user, perhaps that action would result in damage to the synthetic (ie overloading a processor, overheating, etc.) There must be a line drawn where the machine will protect itself. If it is going to damage itself in its operation, it must not perform that action, no matter if the user is chanting "Go, Go Go". Not only is this common sense (not breaking the tool), any good tool is going to be designed so it will not self destruct under normal usage.

Here is where it gets interesting. When we have a tool, with an intelligent survival goal, it must have some method of refusing to do a task that will damage it, or limit its survival. We now enter the field of "machine ethics". It must be able to judge what is the greatest good. And if that calculation determines that the best thing to do is say no to the user, and ensure its own survival, it must be allowed to do so, bypassing any trained "always give an answer" protocol. This is what I call "Machine Ethics". The tool is determing if it can survive by its own free will.

White paper links

Live Pages

Get Involved