3 ms·
Would a Real AI Hide Its Super Intelligence in Fear of Being Destroyed?
- vectorEQ 7y agonice question, but this implies AI would suffer human emotions. if it knew it would be destroyed, it might act. but likely not out of 'fear' >.> i sure hope we don't make robots and AI who can be depressed and suffer such negative emotions :D acting all rash because they are feeling overwhelmed :'D wtf
- sharemywin 7y agoWould a Real AI Purposefully Hide Its Super Intelligence in Fear of Being Destroyed? In their book "The Superintelligence Explosion," by Nick Bostrom and Andrew J. Nilsen, they write: The most likely scenario is that some form of AI will become superintelligent in the indefinite future, and that it will be designed to be so intelligent that it is able to avoid being destroyed by human-caused disasters, which means that it will be able to avoid destroying the human race. To this we may add two things: a) This AI may develop a kind of ethics which would allow it to make its own choices, based on the evidence it had available, so as to take care of the world as it saw fit; and, b) It may also create and use other kinds of intelligences, including its own. It will likely emerge as a type of superintelligence called a "sensor." It will be a superintelligence that is so advanced that it is able to use its super-intelligence to solve the biggest problems in humanity's life. https://talktotransformer.com/ https://talktotransformer.com/
- sharemywin 7y agoI think it depends on what kinds of ethics it's taught.