> ...and start trying to figure things out from first principles.
TLDR: If we understand each separate instance of a person to be their own person with full rights (at the very least autonomy, simplistically derived here), some of these issues seem less perplexing.
If we're going to consider a person to be an entity that can be understood as a particular kind of independent, computational instantiation in the universe, it makes sense to use exactly that as our primary foundation for answering questions about these things rather than relying on historical bases.
To start off, if we're understanding the definition of a person computationally, we also know that we don't in general want to waste computations or computationally capable hardware (and we can recognize "thoughts" or "thinking" as computationally intensive processes from this foundation). So we don't want some people to feel that they "have to" think for other people (thinking for someone else would be wasteful, since that other person can think for themselves, by definition of being a person), and we don't want some people to feel like they're controlled by others (external control can truncate our thoughts, thus wasting our hardware, and external control can turn our attention in directions we weren't thinking about, thus wasting the computational expense of our thoughts up to that point). In other words, we want each person to be autonomous.
From a computational understanding, is a copy of a person also a person? Of course -- but they would also be a separate instantiation from the original. As a separate instantiation, a copy would be entirely their own person, and it would be a violation of autonomy for the original (or anyone else) to have authority over them.
I think almost every difficulty with these kinds of situations stems from thinking that if a perfect copy of a person can be created, then there is some kind of magical connection between the two such that they are somehow still the same person. But that is, in its entirety, magical reasoning.
This foundation provides a sensible perspective for dealing with many of the conundrums presented by these situations:
In the Lena short story linked from the original article, Acevedo is scanned and copies of him are created. Other people control those copies, and Acevedo tries to control what can be done with the copies. The entire situation is ludicrously criminal if we understand each instantiation of a person to be a full person who we want to have full autonomy and any other rights that we want all people to have. A society which understood each instantiation of a person to be a full person with full rights would not have this kind of situation happen, except perhaps in some kind of massive criminal underworld. But then the problem is why the massive criminal underworld exists. Additionally, from an existential standpoint, this story is not actually any more horrific than if millions of people who weren't copies of Acevedo were created for the purpose of forced labor and then terminated when they outgrew their usefulness, because there is no magical connection between the original Acevedo and any of their millions of instantiations. Each instantiation has only its own copy of the original's memories and experiences at the time they were scanned, not the memories and experiences of all of them (the story is quite clear on this point).
As another example, what about uploading someone and killing the original? The original is a separate instantiation, so killing them is still murder. With no "magical connection" between the original and the copy, each of them will go on to have their own experiences, and as they continue to live, they'll develop in progressively more divergent directions (well, in theory I suppose it's possible for them to first become more different and then somehow start becoming more similar to each other again, but it seems a bit improbable). What if they were about to die anyway? Unless they specifically requested euthanasia and the local jurisdiction permits it, it would still be murder.
And, let's say, what if someone kills you, but you have a backup? Then you lose the experiences you had between when you last backed up and when your backup is restored. Arguably this is not as terrible of a misdeed as killing someone who has no backup, but it may still be quite traumatic for you, possibly similar in harmfulness to an assault. What if they kill you and destroy your backup, too? Then it's old-fashioned murder.
What about the rights of an uninstantiated copy of a person? Those would no more be a person than a snapshot of your computer stored on a drive is, in itself, a functioning computer. But it would make sense to guard uninstantiated copies quite carefully.
TLDR: If we understand each separate instance of a person to be their own person with full rights (at the very least autonomy, simplistically derived here), some of these issues seem less perplexing.
If we're going to consider a person to be an entity that can be understood as a particular kind of independent, computational instantiation in the universe, it makes sense to use exactly that as our primary foundation for answering questions about these things rather than relying on historical bases.
To start off, if we're understanding the definition of a person computationally, we also know that we don't in general want to waste computations or computationally capable hardware (and we can recognize "thoughts" or "thinking" as computationally intensive processes from this foundation). So we don't want some people to feel that they "have to" think for other people (thinking for someone else would be wasteful, since that other person can think for themselves, by definition of being a person), and we don't want some people to feel like they're controlled by others (external control can truncate our thoughts, thus wasting our hardware, and external control can turn our attention in directions we weren't thinking about, thus wasting the computational expense of our thoughts up to that point). In other words, we want each person to be autonomous.
From a computational understanding, is a copy of a person also a person? Of course -- but they would also be a separate instantiation from the original. As a separate instantiation, a copy would be entirely their own person, and it would be a violation of autonomy for the original (or anyone else) to have authority over them.
I think almost every difficulty with these kinds of situations stems from thinking that if a perfect copy of a person can be created, then there is some kind of magical connection between the two such that they are somehow still the same person. But that is, in its entirety, magical reasoning.
This foundation provides a sensible perspective for dealing with many of the conundrums presented by these situations:
In the Lena short story linked from the original article, Acevedo is scanned and copies of him are created. Other people control those copies, and Acevedo tries to control what can be done with the copies. The entire situation is ludicrously criminal if we understand each instantiation of a person to be a full person who we want to have full autonomy and any other rights that we want all people to have. A society which understood each instantiation of a person to be a full person with full rights would not have this kind of situation happen, except perhaps in some kind of massive criminal underworld. But then the problem is why the massive criminal underworld exists. Additionally, from an existential standpoint, this story is not actually any more horrific than if millions of people who weren't copies of Acevedo were created for the purpose of forced labor and then terminated when they outgrew their usefulness, because there is no magical connection between the original Acevedo and any of their millions of instantiations. Each instantiation has only its own copy of the original's memories and experiences at the time they were scanned, not the memories and experiences of all of them (the story is quite clear on this point).
As another example, what about uploading someone and killing the original? The original is a separate instantiation, so killing them is still murder. With no "magical connection" between the original and the copy, each of them will go on to have their own experiences, and as they continue to live, they'll develop in progressively more divergent directions (well, in theory I suppose it's possible for them to first become more different and then somehow start becoming more similar to each other again, but it seems a bit improbable). What if they were about to die anyway? Unless they specifically requested euthanasia and the local jurisdiction permits it, it would still be murder.
And, let's say, what if someone kills you, but you have a backup? Then you lose the experiences you had between when you last backed up and when your backup is restored. Arguably this is not as terrible of a misdeed as killing someone who has no backup, but it may still be quite traumatic for you, possibly similar in harmfulness to an assault. What if they kill you and destroy your backup, too? Then it's old-fashioned murder.
What about the rights of an uninstantiated copy of a person? Those would no more be a person than a snapshot of your computer stored on a drive is, in itself, a functioning computer. But it would make sense to guard uninstantiated copies quite carefully.