The top two comments here both suggest that the people using chatbots in place of social interaction were already a priori socially isolated and AI has no effect on it. It's merely a "symptom" or after the fact consequence.
I deeply, fundamentally disagree with that. Humans are one step mathematical operations that take in an input, transform to an output, and are done.
Human life is an endless continuous cascade of incentives, feedback loops, iterations, and modification. When you change anything in a person's environment, it will affect them. Perhaps the effect is small unless someone is primed by their prior environment in certain ways, but nonetheless nearly everything leaves its mark.
Can you eat healthy if your kitchen is full of free junk food? Yes, it's possible. Can you get out of the house and socialize even when endless media and parasocial relationships are just a screen away? Yes, it's possible.
Will you in practice? Evidence shows clearly over and over again that even tiny incentives have huge effects when compounded over time.
We all have a deep moral obligation to build an environment (physical, cultural, social) that is nourishing and incentivizes all of us to flourish. If you're building technology like AI chatbots that enables people to become more socially isolated, in my mind you are in the same category as junk food sellers, drug pushers, and polluting factory owners. You're making people sicker and the world worse.
> Can you get out of the house and socialize even when endless media and parasocial relationships are just a screen away?
The problem is when the junk food is cheaper/more easily available. There are plenty of people I could spend all day with and enjoy it. They are busy living life. The people who have time for me are usually the ones I feel drained after talking to. So I prefer to stay away from them.
Some practical advice I gave a good friend yesterday, who is frustrated by the constant bailing from people to go touch grass with him:
1. Increase frequency of informal communication. For example if your hangout is a monthly coding club, you might casually message participants once or twice a week. “Holy smokes Jenny, this HN thread reminded me of you.”
2. Create convenient little group DMs (call it “Bob’s Coding Club”) and add in the people you like to hang out with.
This way, even if you invite someone privately, this person is well aware you hang out with others; they'll worry you can gossip if they bail too much (even though you won't, but fear of getting ostracized is a typical human trait and therefore a helpful forcing function.)
Source: I run meetups [0] for programmers, many of whom are recovering social media addicts.
Also, if organizing a group thing, get one or two people privately to say that they will go and then announce to the group that the three of you are going and if anybody else wants to come rather than just you are going and to see if anybody wants to come. If there’s already buy in, you’re much more likely to get people to show up.
This is a symptom of alienation, imo, not a unique problem from AI.
There's a theory that we're becoming increasingly removed from our work -- we have less control over what we get to build, we have less control over how the products we build function.
Because we don't influence what we create as much, work becomes much more about getting a pay check. We no longer work to craft, we work simply to build the things the bosses want.
Now that work is just a paycheck, we're increasingly unsettled, and increasingly in competition with one another. Material conditions are such that the bosses get most of the profits, and we get squeezed more and more. Competition gets more desperate, and we begin to see others as threatening our remaining resources, more than a community.
Now that we're increasingly isolated from one another, we end up isolating ourselves. We find ourselves less creative, less fulfilled, more alone, and looking for any semblance of community.
It's not surprising someone in this state turns to anything, even an AI, that wants to engage with the person.
I will argue that work should be just about a paycheck. The CEO/founder's vision is not your vision. It helps you be mindful of differences between your vision and your ethics versus the company's. You'll be able to see more clearly when the gap is too big and it's time to leave.
For me, when work became all about the paycheck, I took my ego out of my work. I remained engaged and performed the job as expected. Whenever I became "unsettled," I took that as a sign to work more on keeping my ego out of my work.
Another advantage of reducing your ego in your work is that you think more about what the customer needs, rather than what the company needs from the product. Doesn't mean you'll make it happen, but at least you know you tried.
To your point about people looking for community, when I reduced my ego at work, I found connection, satisfaction, etc. with communities outside of work.
Even if you're engaged in work, however, you should never lose sight of the fact that the company is making money off the backs of the exploited worker, and it should always remain part of the decision process of stay or go.
This is merely a symptom. The emergence of overwhelmingly powerful and addictive social media apps almost perfectly correlates with this trend of loneliness.
They may not "take people out of society", but they relieve a stress that would otherwise cause them to continue trying to enter society.
Or, you know, go crazy and do something bad.
As a pretty severe introvert, it's not hard for me to imagine "friends apps" getting good enough to provide a lot of comfort when I was younger, and if you grow up with them being a thing, you don't question them as much.
In fact, if "friend apps" were good enough to play video games on a decent level, I'm absolutely sure I'd use them instead of trying to deal with random jerks when playing multiplayer games.
>I honestly can't imagine using an AI app to ever qualify as what I would determine to be what I deem a friend
Agree, but AI has one massive benefit I’ve seen.
Conversations that I can’t have IRL because the response would be “huh? What are you talking about? I don’t know what that is.”
LLM apps are a fountain of knowledge and can reason well enough to bounce ideas and speculation off of. I can dump all my esoteric conversational topics into it and be engaged without needing to do 5 degrees and work in 3 industries at once, and the massive amount of socialization that would entail.
Maybe there will be a division in my social life, where my IRL human connections are more grounded and homely, and the LLM bears the brunt of my neurodivergent interests.
I can see this as I do use it in this fashion. That said I wouldn't ever qualify that kind of call and response as a friendship. It certainly would allow for me to explore topics in depth without trying to filter through networks of people to see who may or may not have an interest there. That said if anything it can scratch interests that I couldn't before.
I'd frame that as a people who want to deep dive can now do that without a huge overhead.
I can't even imagine what is the first step towards exposing yourself to "AI friend apps". You're sitting at home and what? What happens between getting up one day and installing an AI friend?
> What happens between getting up one day and installing an AI friend?
* You get up one day
* You enter a world where fewer and fewer people are trying to form in-person, human connections because they see pseudonymous social media and parasocial relationships as valid substitute
* You try these relationships too and find they are low-effort ways to keep yourself entertained with surface-level human connections
* Forming actual human connections in-person is too high effort (requires leaving the house; has a risk of rejection) so you don't do it
* An AI friend is low-effort and sort of a human connection so you do that
Ultimately I don't see "having an AI friend" as much worse than "'donating' $500 to Asmongold because he's my friend" (note: Asmongold has no idea who you are).
I'm guessing it's a lot of children and lonely teens. They're already used to installing random apps to play around with while having zero awareness or concern over their privacy or how they can be manipulated and taken advantage of by the apps on their devices. I can see it starting out as curiosity or just mindlessly installing whatever the app store advertised to them and it ending up helping to fill a very real void in their lives.
This seems like a good argument for some kind of AI guardrails. And then the question is who decides where the guardrails start and end? I guess the straw man is: if you knew the AI friend was not going to advocate destructive behavior, to oneself (maybe less important?) and to the world around you, then the AI friend might actually be a net benefit because the actual person may feel more emotionally supported than they otherwise would be.
The problem is there's almost no way to regulate the guardrails, at least in a country like the US, and so then you're left with corporate interests deciding on the guardrails with their only incentive being hockey stick growth in engagement/revenue, which we already know is destructive in almost every corporate vertical (social media, food, etc.).
The more robust welfare systems and stronger social safety nets in some European countries buffer against the economic/social stressors that contribute to loneliness.
Japan, however, has other cultural hangups about things like mental health and vulnerability and difference.
I think it's probably a multiple things we need situation -- a robust welfare system and social safety net AND a cultural acceptance of failing and needing that help.
Japan has a super toxic work culture, and economic stagnation *shrug
These problems are like individual health. Everybody knows exactly what the solutions are: Stay active, eat vegetables, get enough sleep, etc etc. Technically simple, but not easy.
Is it? I have my personal anecdote. Most outrageous, fearful and dangerous (for my life and health) interactions with other people I had in public transport.
It sounds like you recognise that they encourage human interaction, though. Learning how to navigate human interaction is very rewarding and it's at the core of what we're losing.
Fundamentally people don't want to be around each other, certainly not strangers, and the perceived (likely even real) consequences of it are an expected net loss.
That's a little extreme. There are country mice and city mice. City mice want to live inside cities and eliminate cars. That's great for them. Country mice, like me and I suspect you, want city mice to stay in the city and not ruin the nicer, quieter living space outside.
Now, if we could just get cities to keep their light pollution within the city, it would be a nicer world.
Readit News
I deeply, fundamentally disagree with that. Humans are one step mathematical operations that take in an input, transform to an output, and are done.
Human life is an endless continuous cascade of incentives, feedback loops, iterations, and modification. When you change anything in a person's environment, it will affect them. Perhaps the effect is small unless someone is primed by their prior environment in certain ways, but nonetheless nearly everything leaves its mark.
Can you eat healthy if your kitchen is full of free junk food? Yes, it's possible. Can you get out of the house and socialize even when endless media and parasocial relationships are just a screen away? Yes, it's possible.
Will you in practice? Evidence shows clearly over and over again that even tiny incentives have huge effects when compounded over time.
We all have a deep moral obligation to build an environment (physical, cultural, social) that is nourishing and incentivizes all of us to flourish. If you're building technology like AI chatbots that enables people to become more socially isolated, in my mind you are in the same category as junk food sellers, drug pushers, and polluting factory owners. You're making people sicker and the world worse.
The problem is when the junk food is cheaper/more easily available. There are plenty of people I could spend all day with and enjoy it. They are busy living life. The people who have time for me are usually the ones I feel drained after talking to. So I prefer to stay away from them.
1. Increase frequency of informal communication. For example if your hangout is a monthly coding club, you might casually message participants once or twice a week. “Holy smokes Jenny, this HN thread reminded me of you.”
2. Create convenient little group DMs (call it “Bob’s Coding Club”) and add in the people you like to hang out with.
This way, even if you invite someone privately, this person is well aware you hang out with others; they'll worry you can gossip if they bail too much (even though you won't, but fear of getting ostracized is a typical human trait and therefore a helpful forcing function.)
Source: I run meetups [0] for programmers, many of whom are recovering social media addicts.
[0] https://handmadecities.com/meetups
It's no surprise that you have excellent advice on how to socialize more. :)
There's a theory that we're becoming increasingly removed from our work -- we have less control over what we get to build, we have less control over how the products we build function.
Because we don't influence what we create as much, work becomes much more about getting a pay check. We no longer work to craft, we work simply to build the things the bosses want.
Now that work is just a paycheck, we're increasingly unsettled, and increasingly in competition with one another. Material conditions are such that the bosses get most of the profits, and we get squeezed more and more. Competition gets more desperate, and we begin to see others as threatening our remaining resources, more than a community.
Now that we're increasingly isolated from one another, we end up isolating ourselves. We find ourselves less creative, less fulfilled, more alone, and looking for any semblance of community.
It's not surprising someone in this state turns to anything, even an AI, that wants to engage with the person.
For me, when work became all about the paycheck, I took my ego out of my work. I remained engaged and performed the job as expected. Whenever I became "unsettled," I took that as a sign to work more on keeping my ego out of my work.
Another advantage of reducing your ego in your work is that you think more about what the customer needs, rather than what the company needs from the product. Doesn't mean you'll make it happen, but at least you know you tried.
To your point about people looking for community, when I reduced my ego at work, I found connection, satisfaction, etc. with communities outside of work.
Even if you're engaged in work, however, you should never lose sight of the fact that the company is making money off the backs of the exploited worker, and it should always remain part of the decision process of stay or go.
If you work somewhere where you have control over the outputs of the your labor, you can both get a paycheck and not be exploited.
But AI exacerbates it, which is the point of the article.
I honestly can't imagine using an AI app to ever qualify as what I would determine to be what I deem a friend.
Or, you know, go crazy and do something bad.
As a pretty severe introvert, it's not hard for me to imagine "friends apps" getting good enough to provide a lot of comfort when I was younger, and if you grow up with them being a thing, you don't question them as much.
In fact, if "friend apps" were good enough to play video games on a decent level, I'm absolutely sure I'd use them instead of trying to deal with random jerks when playing multiplayer games.
Agree, but AI has one massive benefit I’ve seen.
Conversations that I can’t have IRL because the response would be “huh? What are you talking about? I don’t know what that is.”
LLM apps are a fountain of knowledge and can reason well enough to bounce ideas and speculation off of. I can dump all my esoteric conversational topics into it and be engaged without needing to do 5 degrees and work in 3 industries at once, and the massive amount of socialization that would entail.
Maybe there will be a division in my social life, where my IRL human connections are more grounded and homely, and the LLM bears the brunt of my neurodivergent interests.
I'd frame that as a people who want to deep dive can now do that without a huge overhead.
* You get up one day
* You enter a world where fewer and fewer people are trying to form in-person, human connections because they see pseudonymous social media and parasocial relationships as valid substitute
* You try these relationships too and find they are low-effort ways to keep yourself entertained with surface-level human connections
* Forming actual human connections in-person is too high effort (requires leaving the house; has a risk of rejection) so you don't do it
* An AI friend is low-effort and sort of a human connection so you do that
Ultimately I don't see "having an AI friend" as much worse than "'donating' $500 to Asmongold because he's my friend" (note: Asmongold has no idea who you are).
Deleted Comment
The problem is there's almost no way to regulate the guardrails, at least in a country like the US, and so then you're left with corporate interests deciding on the guardrails with their only incentive being hockey stick growth in engagement/revenue, which we already know is destructive in almost every corporate vertical (social media, food, etc.).
https://joint-research-centre.ec.europa.eu/projects-and-acti...
Welfare systems are a net good, but they aren't neccessarily the solution to this problem.
I think it's probably a multiple things we need situation -- a robust welfare system and social safety net AND a cultural acceptance of failing and needing that help.
These problems are like individual health. Everybody knows exactly what the solutions are: Stay active, eat vegetables, get enough sleep, etc etc. Technically simple, but not easy.
Now, if we could just get cities to keep their light pollution within the city, it would be a nicer world.
Speak for yourself. Humans are fundamentally social creatures.
It's the opposite of that. Fundamentally, people do want to be around each other and need to be or they go insane.
There are exceptions, of course, but they don't represent the norm.
Thats why NYC is empty.