True AI(if it ever happens) will have rights. Either granted or taken.
Just because one can declare something does not make it true. One would have to show the machine really is sentient and that's not possible since any complex computing system can mimic sentience and we won't be able to judge it. Also why should a sentient creature be granted special rights? And how many rights to grant it?
How would it take rights? If it takes "rights" then because the machine has such programming it spells nothing for its AI status. Also if it takes rights i.e. dominates us, of course it may excuse itself from responsibility for killing people. It spells nothing about any real moral status it gained after all we agree that might does not make right, right?
But the animals are punished? I just gave you 2 examples(of many).
Animals are not punished and are true AI in the most relevant sense, or equivalents of other forms of intelligence of non human type. They're the closest thing we have to AI. And we punish people. So instead of bullsh-t speculation about robots we can look at how people interact with animals and we don't punish a dolphin which pushes someone overboard and causes them to drown (and that's some complex AI there aleady) or a gorilla which kills someone who fell into its cage.
True AI is NOT programmed, by it's definition... You won't be able to "compel" it to do anything.
Of course you can add programming on top of what the machine executes. You can have circuits which can affect motor function of the machine so for example if a human is recognised as being hurt by a computer which is not-AI the machine is switched off or prevented fron acting, or all its motor inputs go through a computer which determines if the direct action will harm a person or not, and so on.... that is assuming "True AI" is possible. However True AI will still be man made and so can be man controlled to some extent. And if it can't then the people who allow it will be held responsible for any problems it causes. There is no necessity for such AI to continue to exist if it is potentially harmful. So there are many ways people can be held liable. You need to think in a bit of a wider context here.
So how far back does one go in blaming, back to Turing?
No. To the company and people who authorised the use of such a machine. Why Turing? Why not someone earlier? Why not the early mathematicians?
I don't know if it's possible, just incredibly narrow minded to say it's impossible.
It's not narrow minded. There is no evidence that it's possible. Also there is no way to tell if it's really intelligent or mimicking intelligence. Chinese Room problem shows you that people can't judge sentience of something based on interaction with it. So that's evidence against us even knowing it's really, truly intelligent. As we don't even materialistically can account for consciousness we can't know if it's possible for it to truly exist in a non-biological form derived via non-natural means. It may be that we can create exact copies which are philosophical zombies.
So it's maybe possible but to say narrow minded is saying one is narrow minded for being a realist.
True AI even if such a thing can exist won't just pop into existence, well unless we are incredibly lucky with random quantum events, and that is supposedly possible too. It will be developed gradually and won't be able to function as some Terminator robot from Day 1. Those who say such things are safe and release such robots into the wild, will be held responsible for example.
