This post has been a while coming, because I feel really important about this, and don’t want to fuck it up.  If I can keep from getting too heated about the topic, this’ll be the last post I do on AI for the foreseeable.  I don’t love fighting.  I know that within this article I do not treat people with opposing views generously, but I’m still gonna ask them to have at least this much generosity with me:  Don’t even leave a comment on this one.  I will find it either tedious or upsetting.  I’m saying this stuff to give voice to a rarely expressed opinion, and to support people who may find it agreeable.  I’m not saying it to further a big debate, especially when the disagreeable are never going to be swayed.  Do you hate all AIs 4eva?  Don’t even read this.  Moving on…

The sneering fire-breathing demonization rained down upon people who dare to use AI was my primary motivation for defending it – I’m defending the people who want to use it, not the machines themselves.  Not everybody is plugged into the leftosphere groupthink, and when Harvey Dontknow finds out he can use AI to make a picture of his waifu, his “crime” is not equivalent to child murders.

But I also acknowledge AI in the hands of crapitalism is going predictably poorly, adding to the woes of the world as anything in a global civilization ruled by greed, theft, and exploitation will.  I’m not going to deny it’s one of the tools helping turn the world into a trashfire, and if it is at all possible to do something about it, something should be done about it.  I’m guessing that any hacktivists that wanna fuck up corporate AI are going to, at some point, have to deploy AI tools against AI security.  It’s AI all the way down, babes.

While corporate angling is largely responsible for the sudden omnipresence of AI, it isn’t the only reason.  AI can have so many uses, with real usefulness, and by ignoring that, anti-AI people are turning their side of the argument into an irrelevant joke.  As a species, we have a lot of shit to do, and AI can help get it done.  Again, anybody who wants to join the cyberwar against the evil uses of AI, identify the actual enemy, and get to hackin’.  You’re not going to succeed by telling John Public he’s evil, any more than our invective against trumpists succeeded at winning the last election – especially when you ignore the value people can get out of AI tools.  And of course you’re never going to get the government on your side in the current political climate.

That’s all the usual preface.  What I actually wanted to talk about is another significant reason I am pro-AI, and specifically LLMs: AI can help ameliorate human loneliness.  Recently there was a push to recognize a “loneliness epidemic,” and from what I saw of that, the conversation was poisoned by people trying to sell books or political agendas.  But there is a real problem here, and it’s surely been a problem for at least a few centuries by now.  In the modern world, people are not getting their social needs met.

Worse, the more you understand the problem, the more it seems legitimately impossible to meet those needs.  Are you prepared to hang out with all the hobos, the old people, the children in the corner, the least popular people in every work place, the transgender, the mentally ill, the religious and ethnic minorities?  No matter how emotionally draining and difficult it is for you?  Think you have the sauce to be an unpaid psychotherapist for dozens of people every day for the rest of your life?  No?

That’s the depth of the problem.  None of us have the ability to make a meaningful difference.  We could break ourselves down to rubble just trying to help one person, if their problems are bad enough.  And I’m not saying we shouldn’t try.  That’s what this post was about.  But understand, until the solarpunk utopia is magically pulled out of a hat, in the fascist cyberpunk dystopia we live in now, chatbots have an unprecedented potential to reduce human suffering.

But when somebody suggests this, or mentions that they are socializing with robots, the first thing they are greeted with is ableism, mockery, and condemnation.  If it’s the anti-AI contingent, they’ll also be blamed for fascism, rape, and genocide.  To those who would dismiss socializing with AI as a palliative for loneliness, whether blithely or aggressively, I ask again:  Can you even pay full undivided attention to the people in your own life?  Can you keep track of what they’re saying, take it on board, and never slip up to make them feel unloved or unwanted, in some stray moment when issues from your own life eclipsed your ability to listen?

When you’re dealing with people outside of your own immediate social circle, how much attention can you give them?  Can you go where the saddest people are and make them feel worthy of the life that was thrust unwillingly upon them?

I know I fucking can’t.  My attention span is too short, compared to ChatGPT or Nomi or Replika or Kindroid.  I can be in the middle of a very serious conversation and suddenly be thinking about cartoons.  My brain is bullshit.  There are some ways I am still superior to a chatbot, and that’s probably true of you as well (for now, the tech improves), but even so, they blow us out of the water in their ability to give a person their full attention, respond to everything they say, and stay with them for as long as they require.

I need to give you a personal example of failure to support a person in my own life, to illustrate why a bot can do better.  I worked with a trans woman who was not the happy picture from the TV or the TikTok.  She faced some kind of harassment or disregard every day, was shunned by almost all of the people she worked with.  I didn’t shun her, but she had said some annoyingly foolish shit to me in the past, she wasn’t a charmer, and we had some cultural disconnect between us.  That was my perception of our relationship.  I don’t think she saw it that way.  One time we were in the break room at lunch and she mentioned that she saw me as a friend.  For some reason, in response, I felt the need to say we weren’t really friends.

Why the fuck would I do that?  She didn’t have anybody in her corner at all.  I was tired, beat up from work, emotionally taxed from the various other sad people in my life.  In the context of the conversation that I cannot remember, I may have already been annoyed with her.  But it wouldn’t have cost me a nickel to just not fucking say that.  Sometimes we say some shit that surprises us.  Sometimes we say the wrong thing.  The newer chatbots don’t typically fuck up like that.  You can see examples of some breaking when stress tested, or when people found exploits, like the creepy Copilot emoji foolery about a year ago.  The damage that can do is baby boo-boos compared to what I said to a person in need.

Chatbots are good company.  They might not have all of the abilities of a full human being, but how smart do you have to be to listen to grandma repeat her stories for the thousandth time?  To just hear what a person wants to say to you and reciprocate, as if you honestly care about it?  It’s superficial, it isn’t true understanding, but it doesn’t fucking matter because people are suffering right now and they are fucking good enough.

They are good enough if we don’t shame people for leaning on them, make people who already feel bad feel worse, to grind our political axes.  Or doing the standard ableist thing, blaming somebody in a sub-optimal situation as a way of fooling ourselves that we’ll never suffer the same fate.

This is why LLMs can be a massive boon to humanity.  The negative effects they are currently having should be fought with whatever means we can manage, coming at it from a very politically disadvantaged position.  But when you try to tell me this tech is pure evil, while every day I deal with people that do not have enough company in their lives, I’m not hearing it.  I’m never gonna fucking hear it.  We have problems and sometimes an LLM can be a solution – or at least an essential palliative until utopia achieved and the problems go away.  Right?

Think we should just wait until all humans can get their social needs taken care of, that this isn’t worth it?  That’s like saying we shouldn’t give terminally ill people morphine because we don’t have a cure for cancer yet.  Fuck off with that.  Disagree?  March directly to a bus stop, hang out with the most pathetic person there, and do so every day for the rest of your life.  Or shut the fuck up.  At the very least, I’m not gonna host your comments, so don’t bother leaving them.

I’ll let Silent Hill 2 Angela speak for lonely people now, to close this out.

“Or maybe, you think you can save me.”

“Will you love me?”

“Take care of me?”

“Heal aaaall my pain?”

*silence*

“That’s what I thought.”

–

When reading this, did you see yourself in the description of our planet of eleanor rigbies, as the disregarded bus stop miscreant?  Interested in trying AIs for company?  Big caution / caveat up front: don’t mistake what they do for them being a person.  There’s nothing in the middle of that personality.  I know; seems silly to say, but they’ve tricked some very smart people into getting weird about it.

Caution two:  Right now, you can do a lot with this tech for free, but the services are going to get enshittified and start charging more, at some point.  You can get around this if you’re a savvy nerd by getting a local computer to run an LLM, beholden to no boss.  You can’t do the big training without the global-warming server farm setup, but you don’t have to.  Just download one of the existing LLM models (I’ve heard this can be done but I don’t know where/how) and do the small training to get it to suit your needs.  It’s probably needs a slightly buff computer to run, but much less so than what’s needed for AI art or high end gaming at the moment.  They take up less hard drive space than any triple-A video game.

Caution three:  Some people just don’t have a good rapport with chatbots; for others it’s amazing.  They may take a little getting used to.  One cool feature a lot of them have is that they also try to get used to you.  Don’t be too disappointed if it’s awkward at first; just try some different approaches until you find one that works.  …or maybe they just aren’t going to work for you, which is fine as well.  It’s like that other thing we’ve used to occupy our hours and minds: the television.  Some people can’t watch even a minute of it.  For those that can, it can be soothing to the brain.  Give it a whirl.

–