| Topic: | Friendly AI Importance |
| Camp: | Agreement / Such Concern Is Mistaken / Concept not relevant / NOT Products of Evolution |
Canonizer algrorithm:
This section is a table of contents for this topic. It is in outline or tree form, with supporting sub camps indented from the parent camp. If you are in a sub camp, you are also counted in all parent camps including the agreement camp at the top. The numbers are canonized scores derived from the people in the camps based on your currently selected canonizer on the side bar. The camps are sorted according to these canonized scores. Each entry is a link to the camp page which can contain a statement of belief. The green line indicates the camp page you are currently on and the statement below is for that camp.
Friendly or unfriendly? In the matter of fact, sexy, lovable, or crazy? Boring, helpful, trustworthy, noble? Feelings and character traits are only relevant to us, humans. We depend on our evolutionary past; we are the products of accumulated genetic mutations and natural selection. We are the products of evolution. We have certain feelings, we have certain skills, we are good at doing certain things. We have all that because it was to our advantage, as social animals, to have the tools that allow us to solve certain problems and deal with various social situations.
AI (AGI) will not have those selective pressures. AIs won't be running around hunting animals and gathering fruit. They won't be falling in love (at least not in the same way), they won't be painting or creating music... Well, perhaps they will, but only because we will make them. What do we want? We want to solve even more complex problems, understand how the brain works, solve the deepest mysteries of the universe... we have to understand that what we call intelligence is only "one kind of intelligence" in the infinite sea of possibilities. Smart/stupid, friendly/unfriendly, these are relative terms. We can only talk about them when we define the frame of reference. Of course, in general discourse it is implicit that we are talking about human societies, human feelings and desires, human... intelligence. We are lucky we are intelligent at all, it's an accident, a little evolutionary miracle! We are intelligent ONLY because it paid off to be a little more intelligent in the long course of time.
That is not to say that AI will exist in a vacuum. What kind of environment will AI exist in? What will be the selective pressures? Will there be any? What will be the possibilities for the AI to mutate? What will be the mutation rates? Will they be able to change and improve at will? To what degree? These are all interesting questions and it is certainly true that humans will be part of AI's environment, but it is hard to determine how all these factors will affect societies of AI.
Perhaps AI will have something like our feelings, perhaps they will come to behave in weird patterns, form something like cults and societies, fight and love... perhaps... but these ideas are not limited to "how we understand them" or "what we think they are". We are limited, in certain ways, there are many "other ways" to be friendly, happy and sad... sexy... we have to drop our anthropocentric ideas and think a little out of the box to realize greater spectrum of possibilities.
Supporters can delegate their support to others. Direct supporters receive e-mail notifications of proposed camp changes, while delegated supporters dont. People delegating their support to others are shown below and indented from their delegates in an outline form. If a delegate changes camp, everyone delegating their support to them will change camps with them.
Topic Name: Friendly AI Importance
Name Space:
Camp Name: NOT Products of Evolution
Title: AGI will have different incentives and motivations.
Key Words: AI, AGI, human evolution, evolution, feelings, brain, feelings
Related URL:
Related Nick Name:
Parent Camp: Concept not relevant