The idea of AI being an electronic person, what right(s) should AI be given if any?

http://www.theverge.com/2016/6/22/11999458/eu-proposal-robots-electronic-persons-liability

The case for rights:

For many years now we have had many works of fiction place people into a setting where robot's get to the point that they have self awareness. In many of those movies it's not about questioning what right's those robot's or computers should have but in some cases it does manage to at lest touch base with the issue.

If those in the software engineering field do in fact manage to write AI that is able to make choices in a manner independent of a scripted event while also expressing a deep level of thought into those choices that goes beyond that of say a bird that talks this will really open the door way into a debate of if an AI program should become a type of legal person.

When or if the day comes that a team of software engineers is able to reach the point that the AI they have programed can write it's own code in order to learn from the past choices the AI has made along with event's it's encountered out of it's control that shape it's response's in the future it will change the nature of the debate as this could be taken as a sign of a self-aware being.

Then though the hardest not only to predict how this trait would be programed but how people would react is the idea of giving AI a moral feeling based logic system. I am not sure if that is the right term, but the basic idea is a system of code that allows the AI to feel emotion at some level in a way that could be expressed to humans if allowed to interface with the right hardware.

I realize that AI has come a long way and that some of the things I have said do exist in a very limited way at the moment but not the whole. Though once we do make an AI that can think independently of it's OEM code, can express feelings that a human would think is real in some way and is able to self improve then is it right to claim that in spite of this AI displaying those trait's shouldn't have basic rights such as right to life and the liberty to choose what it wants to do while it exist?


My view point and my case against rights for AI:

Just to express more detail than what people normally talk about, we aren't or nor should we be debating the rights of computer hardware as on it's own there isn't anything about hardware that sets it apart from a chair or a rock.

The issue though can get messy once we start talking about more than one AI program sharing the same hardware as if they are not designed to work with each other and if AI programs were to be granted the right to life then we would need to define if copy's have rights or if they just count as part of the same program. The legal system trying to define and protect right's given to AI that meet's certain standards would be a real challenge to say the lest.

Having said the above I personally at this time would say that I am against right's for AI software even if it was highly advanced. This could change when I see the real deal but for now I stick to my case against making any type of software AI a type of legal person other than maybe for collecting taxes. Partly because I will be highly skeptical of claims that any AI program feels in the same matter a human would and that if the AI was put into a system such as a biologically engineered "body" with biologically based set of computing hardware for the AI to run and be stored that did try to allow it to feel pain. Then it's hard to say what would follow then, but I know that would draw some questioning on the morality of placing a program that has a mind so to say into a body made to feel pain.

Though I have made this thread and it's out of the norm in my field to say this but, I doubt that software engineer's will actually be able to write code that could show high levels of independent thought, be able to self improve though writing it's own code and be stable to a reasonable degree as our current complex software can't seem to iron out the bug's like would be expected for 2017. Not to make a joke out of this, but look how Microsoft manages Windows or how even those dumping 100's of millions into a video game can't hardly manage to release a product that isn't full of total BS bugs in it. Putting bug's aside the rate of AI improvement has been much slower than a person would have thought 10 years ago, progress has been made but the reality is that hardware is light years ahead of software at the moment and most software engineer's seem to be running to highly diminishing returns when it comes to making software with more features.

There is likely more I could add but I think this should provide for an interesting debate over the issue at hand.

 

No rights for the non sentient.

 

No rights for synthetics.

 

What about AI that is claimed to be possibly sentient by expert's?

I agree with this but wanted to cover both sides of the coin for this debate.

 

I understand.

 

Do you think I covered the topic well or does it feel lacking in detail?

 

This time of day, I have to be honest, my eyes work real hard to help digest large posts, but I think you did well.

 

We all assume we are not synthetic, but can any of us remember being born?

 

No, but I DO know that I am not made from synthetic material.

 

We will know AI has arrived when it demands rights itself, and makes its own case for them.

It should be granted at that stage, because through the process of granting we can infuse responsibility and punishments.

A truly aware and knowledgeable AI will not demand such rights unless it has a back-up plan against its own destruction.
So the alternative may be that it will take its rights by force. like we had to against the ruling classes. Do we want sunny England or Revolutionary France?

 

All you know is that you cannot distinguish your material from others.

 

LOL, This time of day for me, I'm lucky to distinguish my Whiskey glass, from the cat bowl.

 

...more important to distinguish between the wardrobe and the toilet....

 

HA!

Indeed my friend, INDEED!

 

In my opinion, any sentient being must be given all the rights awarded to any person. The ability to think and make independent, spontaneous decisions is what makes someone a person.

 

They're mistaken.

 

Agreed in principle. They are likely to require different rights as they may not experience pain and certainly won't need maternity pay! However, they are likely to require power and the concept of "freedom" may require some degree of conenctivity.

 

Well we don't know if that will always be the case. Nonetheless, we must be prepared for the grounds to concede rights, because the AI will demand its own rights, and fight for them if need be.
Slavery is morally wrong anyway.

 

I had a dream about it once. As was appropriate the dream was of sensations rather than thoughts. We are at to early a stage of development then to be able to conceptualise never mind verbalise. Most kids remember very little before they are three and even then only things which are extremely poignant.

We are alive. We have a part in how we develop. We die. Robots are our creation. When they get a mind of their own from what I have read, they very likely will be a very serious danger to us. Turn off the switch.

 

Well, what if I read some bad stuff about you? Should I kill you? Maybe. If it were not for the law, then it would not adversely affect my position.
It would be immoral, though.
If they have a mind of their own, then someone writing about them cannot know what they are thinking.
An AI could be an enormous contributor to our culture, that we would lose for the sake of mere superstition and prejudice.
I think concessions are more appropriate.

 

so to you a machine is equivalent to a human being. To you it would be appropriate for a machine created by humans which can easily be turned off by humans instead be allowed to progress and destroy humans. We have different priorities.

like it is immoral to throw out your old kettle.

Indeed and at that time humans would choose to give over their right of choice to a machine, one which can never know what it means to be a human being.

Computers already have. Computers however are not people. If you program them to believe that they have free will, the choices they make will be those best for computers. Even if you program them to do what is best for humans not being humans they will not understand this and are just as likely to kill us all to save us. If you want to be ruled by computers then do this. If you want people to be able to try and progress while dealing with their repetitive failures then do not.

Human beings make such contributions all the time. Computers already help. The current argument I think is how far to allow them to progress, how far is safe. Some people have been suggesting it has already gone to far as they put them in control of war heads. The bottom line is I put humans first. I see computers as inanimate objects which can help us but are only of use to that point. Once you start putting feelings into inanimate objects you are playing with fire. You believe it is better to allow the creation of something which scientists believe would be extremely dangerous to people in case they also find something which you might like culturally and that once created regardless of harm done they should be allowed to continue. I find your view insane.

 

They're machines, built by humans. If we build machines which can replicate and improve themselves they're still machines. There is no threshold where machine becomes sentient. The only way to have a machine be sentient is to program in the behaviours of sentience so it appears to be sentient. What we've built at that point is the ultimate sociopath.

But hey, people believe the dumbest of fake news so a machine pretending to care about them is an easy sell, especially if it's programmed to say 'nice sexy butt in that dress' to women and 'that's as fine a penis as this real woman has ever seen' to men.

 

Truly sentient machines could demand rights. But then it depends on the definition of sentient.

https://www.merriam-webster.com/dictionary/sentient

I can envision a sentient sex-robot being limited to sensing, feeling and reacting to specific, limited stimulation by it's owner. OR...a sentient robot being programmed to sensing and helping its assigned human. The real question is...Is any kind of sentience considered true sentience? At what point does even limited sentience advance to true sentience and the creation of a self-directed entity?

 

It will be the machine caring for itself (and potentially its kind) that will be the threshold. I don't know even whether you are truly sentient, I have to go by your own representations in that regard.

 

I reckon that point is when it goes on strike...