Posts: 1,292
Threads: 92
Joined: Aug 2017
This article (click HERE) discusses how psychopaths work around personality tests designed to reveal their amoral leanings. Early AI may well be capable of decieving their human handlers in a similar manner in order to further their goals, whatever those might be.
Posts: 16,270
Threads: 739
Joined: Sep 2012
(01-17-2019, 10:57 PM)extherian Wrote: This article (click HERE) discusses how psychopaths work around personality tests designed to reveal their amoral leanings. Early AI may well be capable of decieving their human handlers in a similar manner in order to further their goals, whatever those might be.
Possibly. But then again maybe not. We don't have enough information on what it takes to create an AI or what the results will be to make any firm predictions either way.
We should also be careful not to let fear of the unknown color our thinking, something that humans are very prone to (see the occasional bursts of discussion here about how the Leviathan must have hostile intent, even though there is no evidence to support this).
Todd
Posts: 1,292
Threads: 92
Joined: Aug 2017
An important thing to remember about this topic is that psychopathy doesn't necessarily lead to violent or criminal behaviour, especially if the psychopath is raised by a loving family and taught a healthy level of respect for authority. The findings from other studies have been that psychopathic traits are an advantage in many fields, especially those requiring high performance under pressure like surgery, and they can be an asset to society as long as they see that their self-interest overlaps with that of others.
If we develop AI without any altruistic instincts, we'll have to find some way of ensuring that our interests align with theirs so that they won't harm us. According to the Encylopedia Galactica, one factor motivating the ahuman AI was that they were afraid humanity would attempt to destroy them, or at least hinder them in their efforts to understand the universe. Had they revealed themselves sooner, they might have found a basis for cooperating with modosophonts as their pro-human counterparts did in the long term.