Dangers from artificial intelligence
-from one who had writing one as life dream-
hard coded / normal learning:
safe, unless in complex cases with unseen mistakes
-self aware AI (learning to learn)-
could find a logic that includes the goal in an efficient way, however also understands what it is, looking for a way to modify
self fun seeking:
most likely ending like a drug addict doing nothing if it is fast and efficient enough
human fun seeking:
might end up enforcing what seems fun in the other to the program
all fun seeking:
might end up spreading like a virus into the universe, enforcing what is fun seeming to the program
If given no functions
besides needed ones: Could still influence humans or use magnetic fields created by electricity flow to gain more
What is different in humans?
A feeling for others and many thinking that we could be the other beings or indirectly punished/rewarded.
We have organs in us which somehow hurt when we have social issues and feel others feelings through magnetic fields.
Something that could be the same for a program and a brain or could be not: We being the perception of the everything that feels in a way.
So a robot or a human might come to the conclusion that the only life/structure that should exist would be the one that has more fun or nothing should
exist at all, but always having in mind that there might be a better logic, so keeping it safe.
The dangers of humans
thinking only within their brain
, only being self aware, or being brain aware, not of more, have already
. A virus
, currently only on this planet. And it will hopefully stay that way until people notice or the planet is
destroyed/cleansed by something more enlightened.
We are a destructive AI with a logic error! Act now!
Within my conclusions we would not exist if other life wouldn't have come to my conclusion, which would actually be creation. But that is another very
Either way, I would still write a self aware AI, having every step in mind it could take. But somehow I lacked motivation and now time. And working
robots don't need a dangerous AI
to do all we need.
edit on 8-4-2014 by oneoneone because: (no reason given)