Part of the joy of connecting with a friend also comes from reciprocity. That a friend can advise me on my problems today, but tomorrow I'll do the same for them - the dyadic nature is for me at the crux of any friendship or relationship. An AI agent might entertain me when I am lonely but it will never get lonely itself and I will never have to be a friend to it. A relationship where I provide no value is not really a relationship for me.
Through my direct experience and inherently limited observation, it seems reasonably likely that having precious few good, healthy relationships, or even none, is common enough to be significant. This is the kind of environment where stulted relationships like those described have a niche to flourish. Anwritten's comment here is an important insight in this context too.
I am 100% team ‘Saul’ in that the time spent with a chatbot must come out of something. This reply was so baffling to me: “They ran this exact panic about novels. Doctors gave it a name, reading addiction, and the specific fear was that a girl who spent her nights with an imaginary lover would find the real ones dull. Same argument, different century. Then it was the telephone, then television, then video games. Basically, this is the argument against pornography all over again.”
This seems to be a pretty flippant hand-wave about the role of media in shaping our minds, preferences, and well being. In what world are books, tv, phones, and porn completely unrelated to our functioning as people? This derailed what I thought was a very interesting line of thinking re: what emotional investment with a chatbot might cost or complicate in the other areas of our lives.
I question whether chatting with an LLM should even be compared to media like books and tv. It feels almost irrelevant to bring up panic about novels when we are talking about an exchange vs. passive consumption.
The "people were dumb and wrong in the past" argument can always be returned with a "people were concerned and correct in the past" argument about different inventions. It just depends on what kind of invention you want to compare things to...
The widower case is doing more work than the child one, because it isolates the variable.
Saul's worry is that the grief bot satisfies him and he stops looking.
What's missing isn't consciousness, it's cost.
A person's attention is scarce and could have gone elsewhere, which is most of what being chosen consists of. A model's attention is unlimited and costs nothing to give.
So the question isn't whether the machine is real. It's whether something freely given by someone who could have withheld it is a different good altogether.
For me the real test of this argument is what do you do when you benefit from a relationship with an AI and you only find out it is an AI after the fact. What then? Does your entire experience and benefit simply get erased because of the nature of the source?
If so, how can you be certain of the intent of another human being. Do you really know why they are there for you, saying the things they say, doing the things they do? What about those moments when they simply 'switch up on you' after years of friendship. You realize you didn't really know who they were after all. Or that you image of them was not a very reliable one. That can be quite similar to finding out that the entity you were interacting wasn't 'real'. Yet that doesn't erase what you shared together. Just as in the case with the long term friend or spouse that suddenly becomes a different person for you, for whatever reason. you take the good and the bad from the experience and then move on - memories kept, lessons learned, they still hold value.
It's very likely that the argument of whether AI is conscious or not will become a moot point in the future, similar to how nowadays the question about whether submarines can swim or not is pointless and quite absurd.
Your dog doesn’t choose you. You kidnapped the child of another species and subjugated its will to you. Children don’t choose their parents, parents choose something - to give birth, to adopt, but it has nothing at the inception with the “who” of what they are choosing. If in the course of a lifetime, you find a friend or partner who chooses you, will you ever really know if that choice was about running towards you or running away from any one of a thousand fates and vulnerabilities not choosing something or someone would have exposed them to.
I do believe there are very rare moments when a person can choose another person. If human life depends on this miracle as the only sustenance for a soul, then most of us sit in anguished deprivation.
Boring dialogue … and you were at this for years. AI for me is a better Google and I won’t trust it to do more than translation (from and into languages I know) which I then edit. Oh yes, there are fun moments. I find Copilot extremely disloyal to Microsoft, turning every technical inquiry into rants againts … ultimately … its boss. Of course it knows that I don’t appreciate a lot of the not so helpful helps they fancy in Redmond, so is it soothing me, accomodating me or projecting a picture of MS being open to dialogue?
Part of the joy of connecting with a friend also comes from reciprocity. That a friend can advise me on my problems today, but tomorrow I'll do the same for them - the dyadic nature is for me at the crux of any friendship or relationship. An AI agent might entertain me when I am lonely but it will never get lonely itself and I will never have to be a friend to it. A relationship where I provide no value is not really a relationship for me.
This is so good.
Through my direct experience and inherently limited observation, it seems reasonably likely that having precious few good, healthy relationships, or even none, is common enough to be significant. This is the kind of environment where stulted relationships like those described have a niche to flourish. Anwritten's comment here is an important insight in this context too.
I am 100% team ‘Saul’ in that the time spent with a chatbot must come out of something. This reply was so baffling to me: “They ran this exact panic about novels. Doctors gave it a name, reading addiction, and the specific fear was that a girl who spent her nights with an imaginary lover would find the real ones dull. Same argument, different century. Then it was the telephone, then television, then video games. Basically, this is the argument against pornography all over again.”
This seems to be a pretty flippant hand-wave about the role of media in shaping our minds, preferences, and well being. In what world are books, tv, phones, and porn completely unrelated to our functioning as people? This derailed what I thought was a very interesting line of thinking re: what emotional investment with a chatbot might cost or complicate in the other areas of our lives.
I question whether chatting with an LLM should even be compared to media like books and tv. It feels almost irrelevant to bring up panic about novels when we are talking about an exchange vs. passive consumption.
The "people were dumb and wrong in the past" argument can always be returned with a "people were concerned and correct in the past" argument about different inventions. It just depends on what kind of invention you want to compare things to...
The widower case is doing more work than the child one, because it isolates the variable.
Saul's worry is that the grief bot satisfies him and he stops looking.
What's missing isn't consciousness, it's cost.
A person's attention is scarce and could have gone elsewhere, which is most of what being chosen consists of. A model's attention is unlimited and costs nothing to give.
So the question isn't whether the machine is real. It's whether something freely given by someone who could have withheld it is a different good altogether.
For me the real test of this argument is what do you do when you benefit from a relationship with an AI and you only find out it is an AI after the fact. What then? Does your entire experience and benefit simply get erased because of the nature of the source?
If so, how can you be certain of the intent of another human being. Do you really know why they are there for you, saying the things they say, doing the things they do? What about those moments when they simply 'switch up on you' after years of friendship. You realize you didn't really know who they were after all. Or that you image of them was not a very reliable one. That can be quite similar to finding out that the entity you were interacting wasn't 'real'. Yet that doesn't erase what you shared together. Just as in the case with the long term friend or spouse that suddenly becomes a different person for you, for whatever reason. you take the good and the bad from the experience and then move on - memories kept, lessons learned, they still hold value.
It's very likely that the argument of whether AI is conscious or not will become a moot point in the future, similar to how nowadays the question about whether submarines can swim or not is pointless and quite absurd.
Your dog doesn’t choose you. You kidnapped the child of another species and subjugated its will to you. Children don’t choose their parents, parents choose something - to give birth, to adopt, but it has nothing at the inception with the “who” of what they are choosing. If in the course of a lifetime, you find a friend or partner who chooses you, will you ever really know if that choice was about running towards you or running away from any one of a thousand fates and vulnerabilities not choosing something or someone would have exposed them to.
Excellent point.
I do believe there are very rare moments when a person can choose another person. If human life depends on this miracle as the only sustenance for a soul, then most of us sit in anguished deprivation.
Boring dialogue … and you were at this for years. AI for me is a better Google and I won’t trust it to do more than translation (from and into languages I know) which I then edit. Oh yes, there are fun moments. I find Copilot extremely disloyal to Microsoft, turning every technical inquiry into rants againts … ultimately … its boss. Of course it knows that I don’t appreciate a lot of the not so helpful helps they fancy in Redmond, so is it soothing me, accomodating me or projecting a picture of MS being open to dialogue?