3 ms·
> We already know the answer to this: anthropomorphization is a near universal cognitive bias, by which humans assume a mind behind phenomena, particularly phen
by vidarh 4y ago
> We already know the answer to this: anthropomorphization is a near universal cognitive bias, by which humans assume a mind behind phenomena, particularly phenomena that exhibits complex behvaiour. This is why people are polite when conversing with chatbots like ChatGPT.
That does not give us the answer to what I wrote at all. Yes, people will casually treat computer systems as thinking entities. I wrote my first chatbot that fooled people ~28 years ago or so. It was not a good chatbot: All it took to fool people into long conversations was to continuously pick random responses from a set of only 4-5 or so.
But that does not tell us whether or not they would think that a computer exhibits free will, nor whether they think it ever could. Assuming this cognitive bias equals a belief that is maintained when probed deeper is something I've seen many times when discussing this with compatibilists. It's easy to assume a compatibilist view on free will if you look at surface level indications; people are not consistent, nor do most think carefully through unstated implications. We - me included - will often act as if something has agency and/or is a fully thinking being with a meaningful form "magical" free will even when we know full well the entity in question isn't even an advanced automaton.
Case in point: the trivial chatbot mentioned above. Some people would pour out their life story. Some would get angry at it for repeating itself. Some would tell it they thought it was probably a bot, and still talk to it as if it was a sentient being afterwards. Does that mean they have thought it had agency anyway since they treated it as if it did after "outing" it as a bot? Of course not.
It meant that in that context it met some immediate desire or need of theirs to treat it as if it was another mind they could connect with, because it was a convenient fiction.
Compatibilist free will is a convenient fiction. And so if you were to ask me whether I agree that "free will" as defined by compatibilists exists, I'd say yes. I just don't find it to have any moral implications, and I see it as no more than an illusion and a smokescreen. And so if you were to ask whether I think "free will" exists, I say no. It's an absurd notion. There's nothing "free" about the compatibilist notion of free will.
Failing to account for this ability to buy into a fiction like this on a contextual basis makes any investigation into what people actually think of something like free will entirely meaningless twaddle (and to be clear, most of what I've seen from non-compatibilists in this respect is also meaningless twaddle). Trying to dig into this with short surveys without multiple questions aimed at determining inconsistencies in the answers that signal lack of understanding of implications is a big red flag.
> If there's a mind that's responsive to moral feedback as I described above, then it makes perfect sense to hold it responsible, and I am 99.9999% certain this is how most people would treat this question.
Even if we posit that people would treat it that way, it does not mean it makes perfect sense or is justified. If that what people do believe, it would just tell me that most people are just as immoral, cruel and callous as they consider criminals to be. Personally, if one consider it moral to apply a corrective system to attempt to reduce crime (and I do, at least for crime that actually has victims), then I'd consider it equally moral to apply the same form of corrections to people engaged in that harm against others as part of current "justice" systems. Participating in the oppression of the huge number of people being punished in ways that do not serve a corrective purpose the way prosecutors, judges and police do amounts to horrifying levels of immorality from the point of view that the people they inflict harm on could not have acted differently.