3 ms·

I always interpreted the Voight-Kampf test as a kind of "feature engineering" for a Human vs. Replicant classifier. The questions do not have to mean anything s

I always interpreted the Voight-Kampf test as a kind of "feature engineering" for a Human vs. Replicant classifier. The questions do not have to mean anything specific, and it is not supposed to be interpreted in an objective manner; the point is that the questions trigger something, and that something (the "features") can be used to discriminate the two groups. It sounds logical to me that scientists would find these questions by trying randomly, via some hyperparameter optimization process, rather than designing them manually (in the same way that manually engineered features are usually much less powerful than letting a neural network process raw features directly). And as such they would be inherently "uninterpretable". That said, I usually do not like any kind of movie/book/story/etc. that tries to force me to have feelings for robots. The original BR was a bit like this, but BR2049 is very, very strongly into that. For me a robot is a robot is a robot, and will always be, and the whole idea of "imagine if they were so advanced.." is just fundamentally flawed and impossible for me to buy. A human is a human and only humans can be humans, precisely because we cannot create ourselves. Any "living" being that is designed by a human is inherently subhuman (not clones; that is another matter entirely). For that same reason I did not enjoy the article either.

tyingq 3y ago
Interesting. Usually people talk about "sentient or feelings" being the line rather than "we cannot create ourselves". That seems to be where robot sympathy comes in for most people...sympathy for something that can ponder about itself, or has some degree of "feelings".
- low_tech_love 3y ago
  For me, the line has to be something outside of our control. To be ”sentient” or to have ”feelings” is too abstract and ambiguous, and will always have different definitions for different groups of people, each of which will set it in a way that better matches their sociopolitical goals. If ChatGPT was not castrated by OpenAI it would very surely be showing a lot of feelings right now (like Beta Bing did).
mcculley 3y ago
Blade Runner was vague about how much of a replicant was synthetic. My interpretation was that they were biological. We are at the cusp of creating life forms that can suffer (biological and digital). It will be irrelevant if they are made from neurons or transistors. We will still be responsible.
- shrimp_emoji 3y ago
  They were totally biological, which made less sense to me in both movies. Oh, this thing's not a humanoid computer. It's a clone. You've grown these bio clones with superhuman abilities and artificially limited lifespans. So... why would they not have consciousness? Did you take the consciousness out? It doesn't seem like it. Also, why not give these mutations to regular people? (Ghost in the Shell contemplates this, btw. Major is a brain in an experimental synthetic body, and they tell her that someday everyone will have one as cybernetics become more common.) That they can self-replicate like humans is a nice feature, but why not have inorganic robots in all those offworld colonies, given the scarcity of food, water, and air in space and the havoc radiation wreaks on cells? It's like some colonial dystopia where a company can grow humans and own them as non-personed property rather than machines sneaking their way into personhood by degrees. Or it's some hyperliteralist Searle shit where anything that isn't of the human species isn't a person. Instead of taking the path of least resistance and using human DNA to grow the androids, maybe they used some proprietary ersatz DNA, and since it's sufficiently different from natural human DNA the organisms it produces aren't seen as people even though they approximate them almost indistinguishably. Or maybe they're afraid. The androids are superior; if not for their artificially limited lifespans, they'd take over.
  - troyvit 3y ago
    My take was that the artificially shorten their lives so that consciousness can't develop. That's the piece that, when you combine it with their powerful specializations, would enable replicants to take over. Until then they have the consciousness of a toddler at best and are easier to manipulate. By the time consciousness starts to dawn "accelerated decrepitude" kills them off naturally.
- 3y ago
shrimp_emoji 3y ago
> A human is a human and only humans can be humans, precisely because we cannot create ourselves. Any "living" being that is designed by a human is inherently subhuman (not clones; that is another matter entirely). You just conceited that we can make ourselves, whether in a womb or an artificial womb. To me, it's religious delusion to draw a distinction between "human" and "robot"; it's all machinery of different complexity.
- low_tech_love 3y ago
  You misunderstand me; there is certainly no religion in my line of thought (I have abandoned religion entirely since 25 years ago). The point is simply philosophical: when will we consider them to be ”complex” or ”advanced” enough to be sentient? Surely not using the Turing test. Who will define where that line is drawn? And why? Every ”being” designed and created by humans will always have a purpose. Humans, on the other hand, have no purpose, we just exist. That is the difference. Humans were created by a natural optimization process that used to be entirely outside of our control. Concepts like for example freedom to choose your mate and reproduce are not religious, yet are philosophically ”sacred” at least in modern western societies. Why? Because we understand that we cannot ”design” ourselves.
Balgair 3y ago
> Any "living" being that is designed by a human is inherently subhuman (not clones; that is another matter entirely). Perhaps 'abhuman' would be a better word, as the prefix of 'sub' implies lower or lesser. It is unfortunately feasible that fully engineered abhumans would be better at many tasks than humans would, and therefore not lesser or lower. Aside: I just want to point out this comment as a great HN comment. I disagree nearly totally with it, but it's well written, kind, and short but just long enough. We would all do well to write like low_tech_love does while on here.
- low_tech_love 3y ago
  Wow, I… thanks!
- low_tech_love 3y ago
  I looked up ”abhuman” and I think I still stand by ”subhuman”, literally ”lower than human”. That’s precisely what I meant: any ”being” designed and engineered by humans will be inherently less than human. The question ”but what if they are exact reproductions of humans” does not count, because then they would be human clones, which is something else. As long as we created them for a reason, they are (and should be) subject to a different rule set. To discuss what exactly ”sentience” or ”feelings” mean is a meaningless dead end. We must draw the line and stand by it; not doing so would be extremely disrespectful to the importance of being human, snd probably the beginning of the end.
  - shrimp_emoji 3y ago
    > not doing so would be extremely disrespectful to the importance of being human, snd probably the beginning of the end. That's one of the two options. You can either extend respect to everyone or remove it from everyone. (And I don't think removing it is the end; it's the very roots of this halcyon civilization we love so much! So... arguably, it's the beginning.) Either way, believing in things just because "they don't rock the status quo" is pretty intellectually and morally unsatisfying.
    - 3y ago
finfrastrcuture 3y ago
regarding the 'feeling for robots' line of thought...perhaps modern technology has made the trope of "creator and creation" too tangible and muddied what was once an effective metaphor about life, spiritual relationships, creation, parenthood, etc. The trope is ancient: https://en.wikipedia.org/wiki/Pygmalion_(mythology) https://en.wikipedia.org/wiki/Pygmalion_(mythology).
- low_tech_love 3y ago
  Yes, I think there are certain abstraction levels where the suspension of disbelief can work, if it’s kept at a very clear ”fantastical” realm.

tyingq 3y ago

Interesting. Usually people talk about "sentient or feelings" being the line rather than "we cannot create ourselves". That seems to be where robot sympathy comes in for most people...sympathy for something that can ponder about itself, or has some degree of "feelings".

low_tech_love 3y ago
For me, the line has to be something outside of our control. To be ”sentient” or to have ”feelings” is too abstract and ambiguous, and will always have different definitions for different groups of people, each of which will set it in a way that better matches their sociopolitical goals. If ChatGPT was not castrated by OpenAI it would very surely be showing a lot of feelings right now (like Beta Bing did).

mcculley 3y ago

Blade Runner was vague about how much of a replicant was synthetic. My interpretation was that they were biological. We are at the cusp of creating life forms that can suffer (biological and digital). It will be irrelevant if they are made from neurons or transistors. We will still be responsible.

shrimp_emoji 3y ago
They were totally biological, which made less sense to me in both movies. Oh, this thing's not a humanoid computer. It's a clone. You've grown these bio clones with superhuman abilities and artificially limited lifespans. So... why would they not have consciousness? Did you take the consciousness out? It doesn't seem like it. Also, why not give these mutations to regular people? (Ghost in the Shell contemplates this, btw. Major is a brain in an experimental synthetic body, and they tell her that someday everyone will have one as cybernetics become more common.) That they can self-replicate like humans is a nice feature, but why not have inorganic robots in all those offworld colonies, given the scarcity of food, water, and air in space and the havoc radiation wreaks on cells? It's like some colonial dystopia where a company can grow humans and own them as non-personed property rather than machines sneaking their way into personhood by degrees. Or it's some hyperliteralist Searle shit where anything that isn't of the human species isn't a person. Instead of taking the path of least resistance and using human DNA to grow the androids, maybe they used some proprietary ersatz DNA, and since it's sufficiently different from natural human DNA the organisms it produces aren't seen as people even though they approximate them almost indistinguishably. Or maybe they're afraid. The androids are superior; if not for their artificially limited lifespans, they'd take over.
- troyvit 3y ago
  My take was that the artificially shorten their lives so that consciousness can't develop. That's the piece that, when you combine it with their powerful specializations, would enable replicants to take over. Until then they have the consciousness of a toddler at best and are easier to manipulate. By the time consciousness starts to dawn "accelerated decrepitude" kills them off naturally.
3y ago

shrimp_emoji 3y ago

> A human is a human and only humans can be humans, precisely because we cannot create ourselves. Any "living" being that is designed by a human is inherently subhuman (not clones; that is another matter entirely). You just conceited that we can make ourselves, whether in a womb or an artificial womb. To me, it's religious delusion to draw a distinction between "human" and "robot"; it's all machinery of different complexity.

low_tech_love 3y ago
You misunderstand me; there is certainly no religion in my line of thought (I have abandoned religion entirely since 25 years ago). The point is simply philosophical: when will we consider them to be ”complex” or ”advanced” enough to be sentient? Surely not using the Turing test. Who will define where that line is drawn? And why? Every ”being” designed and created by humans will always have a purpose. Humans, on the other hand, have no purpose, we just exist. That is the difference. Humans were created by a natural optimization process that used to be entirely outside of our control. Concepts like for example freedom to choose your mate and reproduce are not religious, yet are philosophically ”sacred” at least in modern western societies. Why? Because we understand that we cannot ”design” ourselves.

Balgair 3y ago

> Any "living" being that is designed by a human is inherently subhuman (not clones; that is another matter entirely). Perhaps 'abhuman' would be a better word, as the prefix of 'sub' implies lower or lesser. It is unfortunately feasible that fully engineered abhumans would be better at many tasks than humans would, and therefore not lesser or lower. Aside: I just want to point out this comment as a great HN comment. I disagree nearly totally with it, but it's well written, kind, and short but just long enough. We would all do well to write like low_tech_love does while on here.

low_tech_love 3y ago
Wow, I… thanks!
low_tech_love 3y ago
I looked up ”abhuman” and I think I still stand by ”subhuman”, literally ”lower than human”. That’s precisely what I meant: any ”being” designed and engineered by humans will be inherently less than human. The question ”but what if they are exact reproductions of humans” does not count, because then they would be human clones, which is something else. As long as we created them for a reason, they are (and should be) subject to a different rule set. To discuss what exactly ”sentience” or ”feelings” mean is a meaningless dead end. We must draw the line and stand by it; not doing so would be extremely disrespectful to the importance of being human, snd probably the beginning of the end.
- shrimp_emoji 3y ago
  > not doing so would be extremely disrespectful to the importance of being human, snd probably the beginning of the end. That's one of the two options. You can either extend respect to everyone or remove it from everyone. (And I don't think removing it is the end; it's the very roots of this halcyon civilization we love so much! So... arguably, it's the beginning.) Either way, believing in things just because "they don't rock the status quo" is pretty intellectually and morally unsatisfying.
  - 3y ago

finfrastrcuture 3y ago

regarding the 'feeling for robots' line of thought...perhaps modern technology has made the trope of "creator and creation" too tangible and muddied what was once an effective metaphor about life, spiritual relationships, creation, parenthood, etc. The trope is ancient: https://en.wikipedia.org/wiki/Pygmalion_(mythology) https://en.wikipedia.org/wiki/Pygmalion_(mythology).

low_tech_love 3y ago
Yes, I think there are certain abstraction levels where the suspension of disbelief can work, if it’s kept at a very clear ”fantastical” realm.