In response to Watcher's post:
You are looking at AI from very laic point of view, making the AI like us, having the same values and such, but not treating it like a slave but more of a essential part of society, though worker, is one of the safest ways to avoid machine rebellion, as the AI would essentially want to be treated as it's "living" fellows and work for the betterment of society
Giving AI a simple, even though precise, goal is one of the easiest way to achieve said rebellion, as the AI will soon determine that humans/tau are not needed to achieve that goal, and will likely hold it back, either through moral values or simply being resources better used elsewhere to achieve its goal, for example, individual tau are not required to live in order to protect the tau empire, and would be best used as slaves in the better setting, but more so as bait, desease carriers, being threat to the empire, or simply not being needed, and thus needed to be rid of in order to optimise the efficiency of achieving given goal
Also, the tau by far do not have what it needs to program true AI, so as usual, the AI would need to be achieved by self learning, which is dangerous road, as there is no way of enforcing compliance, and the AI would get too advanced for the tau to understand it and or shut it down mere weeks after it would reach the inteligence level of normal Tau
The Asimov's laws of robotics, amazing as they might be, are in no way apliable to full AI, as it will find the way to get around those, not from malice or evil intend of any kind, but for the very purpose to achieve its need to perform it's ultimate task more efficiently, and we didn't even takwn in consideration the effect of Ruinous Powers, and if you think that wouldn't apply on AI, there is already an example of chaos corrupting men of steel in one of the Gaunt's Ghosts novels
Decent example of such premise would be amazing, but largely unknown movie "automata"
but if you really want to learn about the problematics of AI, I suggest checking out Isaac Arthur's youtube channel, the guy really knows his onions, and his channel is by far the best channel about science and futurism stuff I have seen, though you might have trouble to understand his speech for first few episodes, you can turn on the captions until you get used to it, and I suggest checking his videos out to everyone in this thread, as you are likely interested in the content he is making since we are talking about futurism, transhumanism and AI here
I hope this post makes sense to you, and doesn't come off negative, as I am pretty sleepy right now so I might have trouble to write what I want in a way I would want
Last edited by Beerson
on Feb 07 2018 05:42, edited 1 time in total.