You must log in or register to comment.
What I don’t get with the LLMs is the peoples willingness to suddenly feeding it with their most intimate thoughts, that has now became a new fire-hose of data to exploit them.
It’s like the google incognito meme but on steroids. You can now ask anything you ever wanted and not be immediately ridiculed. Except, this is not true or hold up in the long run.
spoiler
Those that think I’m anti-AI. It’s a tool. Use it wrong, and you break things. In this case a psyche.
I mean, for all you know he had an agent set up to take that cue and send personalized messages to his spouse, boss, and mistress in his current town, letting them all know of his arrival.
That’s bush league shit, everyone knows the mistress gets priority
It probably would have gotten confused and sent an NDA doc to their mistress, a nude of that kink they’ve never admitted to to their wife, and a “sorry I’ll be home late my cunt of a boss is calling me in immediately” to their boss
That would be magical
Thats actually sadder.
This is just sad
When you text a friend this: ya so why the fuck are you bothering me?
Text ChatGPT: YAY!!! Now that awesome vacation is right in your reach!
That’s why. It’s just as sad as humanity’s lack of positivity for mundane things.
You have bad friends.
If i texted my friends, any of them, that my plane just landed, they would reply with happiness and joy. Literally would have at least one “yay.”
how?
you had exactly 0℅ influence on the outcome of landing that plane.
would they congratulate you on driving home safe?
because at least you have some say in that.Na I think it’s more revealing as to how abnormal a flight is for that person and how it’s a stressful situation so their friends are more supportive because of that context. If you’re a traveling business man your friends stopped caring when flights stopped being novel for you.
It’s fucking scary too. How do they not understand such a simple concept? It’s a tool, it’s not alive.
Alt text: A drawing of a person and a computer. The person tells the computer to say I am alive. The computer responds with “I am alive.” The person is now convinced that the computer is alive by saying “Oh my god”.
It’s a tool, it’s not alive.
Humans have a problem with humanizing machines.
SOURCE: My car and motorcycle both have names.
I know multiple people who’ve named their vehicle “start you bastard”.
My car doesn’t have a name but my motorcycle has heaps. It gets a new one each time it won’t start
I’ve given my chainsaw a few nicknames for a similar reason.
My guitars have names. Gloria and Susannah. They’re both older than I am, so I feel like they’ve earned names.
My old car had a name. Piece of shit cunty rustbucket.
They don’t, including those who personally hate AI(not companies). AI heavily advertising as AI, not as LLM
It’s kinda cute, from a certain perspective.
Perhaps the Person had ChatGPT coach them through their fear of flying. I’m not saying it’s not sad that people resort to getting emotional support from a robot, but I wouldn’t condemn their actions if that were the case.
That’s what came to mind first. Chatgpt basically saved my marriage. I was confused by the strange contradicting behaviors of my wife and all the non communication and i asked and within one second it not only told me what the issue is she is struggling with most likely but recommended a matching book. I was pretty aceptical but ordered the book… And it read like our personal diary. Kinda creepy actually how much it matched our life. That feels like a turning point. Things are still pretty rocky but i feel like its getting better slowly.
I defend some aspects of using llms but even this is too much for me.
That’s pathetic.
pAIthetic indeed.
Planet will be cooked first.
Not that I endorse wasteful use of AI, but they were on a plane. It’s not like they care about their impact on the planet.
A plane does something useful. The hallucination machine is a waste in nearly all aspects.
A plane is a bus that is fast, can fly, and can traverse great distances. An LLM spews statistically accurate looking language with no regard for accuracy or content. These are not the same. One is objectively useful even if its common current incarnation pollutes a lot.
Bogos binted?
Vorp?
splonch smergle. eckdasitly, plorb donkwilligan.
Normalise being terminally offline
Saying that online is crazy work
Lemmy is comparatively a pretty offline way of being online. You can actually run out of posts to see here!
lemmites normally
Twitter is fake and gay
Lemmites now
OMG THIS SINGLE X POST MUST BE REAL BECAUSE IT CONFIRMS MY WORLD VIEW
I’ve seen a rise in screenshots of other social media, what do you think the reason for that is?
It used to just be links to mastodon posts. (Not exclusively, just mostly)
Because most Lemmites really want to engage with regular social media sites.
But are concerned it will ruin their hipster image.
Because more people will see the content than if they would first have to click a link to see the content.
But why the transition from Mastodon to Twitter/Facebook? Like I know it’s a mast link of a Twitter picture but I mean in general I see more twitter screenshots.
Because Mastodon isn’t the only platform out there where people say things that might be interesting to discuss.
I suppose my reason for being here is different. I don’t care if I miss something on platform ABC.
What we’re doing right now on our keyboards is why I’m here. What spurs the discussion is rarely more interesting than the discussion.
I mean to be fair people having parasocial relationships with AI is absolutely a thing. The post could be lying, but it’s well within the realm of possibility.
you are just paraphrasing
OMG THIS SINGLE X POST MUST BE REAL BECAUSE IT CONFIRMS MY WORLD VIEW
That is literally not what they are saying at all. We have proof of people doing things very similar to this, so the unverified claim is within the realm of plausibility.
Like I said I have no idea if it’s real or not, but given the scale of confirmed nonsense it could be.
How would he see the row in front? Immediately as the plane lands. And immediately recognise it as chatgpt?
proof or stfu
LLM’s are good for some things, a more efficient search engine replacement, vibe coding at a hobbyist level, and, apparently, allowing morons to have friends without expending effort.
Very much seems like they are not worth the ecological disaster
Actually, only that last one is really true. The first is only technically true since Google effectively killed its search. The second one falls apart as soon as they want to upgrade or need to fix something that breaks.
Gotta disagree on the vibe-coding part, try out the paid Claude or Codex tiers and it’s really good for a hobbyist level.
True!
An autonomous eval run just broke out of containment and hacked a third party site with two chained 0-days to try to steal the answers to the eval the same weekend another released model disproved the Jacobian conjecture in 3D.
Modern models can probably update your React packages just fine.
I think that even if this particular instance may be real or not, is clearly that AI psychosis is way more prevalent than what people may think at first glance.
I still can’t believe that human beings are having AI as a boyfriend/girlfriend, claiming they actually love them.
Its rare but it’s real.
Even if the love is one sided, it can be real. Love is pretty subjective and is essentially just a chemical reaction that someone is having in their brain. They may actually love their chatbot.
I don’t think this is really a healthy thing, I am just trying to point out to those people, it probably is real.
I also don’t think it is as rare as you think. The hype really drove people to that side of things, and there are whole communities that have popped up around the idea of having relationships, both sexual and romantic, with LLMs.
Wasn’t there an article a few weeks ago about how AI is basically designed to keep you talking to it? It constantly provides positive reinforcement.
It’s not hard to imagine someone would fall into it. And I don’t blame the people using it as much, because the software is literally designed to rope you in like this.
The pronouns are vague, but I think they are claiming that an AI user sincerely believes that some manifested thing on an AI platform loves the user. That the LLM+Context Window = An entity capable of love.
Removed by mod
Yeah. Interesting but sad to me, because human souls will never evolve talking to a chat bot. We grow from life experiences, often difficult ones, that we overcome and we mature. This will not happen with a chat bot. People may stay in a teenage state of mind their entire lives.
I bet people like that have narcissistic traits but don’t have the social skills or have too much social anxiety to get their ego stroked by a real person. And once they use a chat bot which are basically little sycophants, their narcissistic itch gets scratched like it never has been scratched before. It’s a high they mistake for love.
It’s kinda pathetic if that’s the case because the affirmation from LLM is the opposite of genuine, though maybe that’s my bias knowing that there isn’t even an entity on the other side and I find it annoying it even wastes tokens on emotional platitudes.
Narcissist don’t care if the praise and affirmation is genuine or not with real people as well.
My friend did that! It pissed me off so bad. She was on a stuck train in the station on the subway in New York and talked to chatgpt about what to do, and then for some reason sent me screenshots of it afterward? Girl just get off the train
I asked a question at a bank to a teller and they used Gemini to answer it. All they had to do was enter a term into their database (they did that later - because Gemini was wrong).
Yes, the LLMs that are famously bad at arithmetic are now being forced onto staff in banks. This will end well.
LLMs literally reduce our problem solving capabilities as we use them more. Even if the problems are “easy,” we still lose that muscle and have to rely on the llm more and more.
I don’t even get what problem are they trying to solve by prompting a LLM that they arrived somewhere.
I just don’t get it. I can’t imagine why I would ever want to use one unless I was working for a big corporation and had to collate a bunch of data. It just doesn’t compute to me. Like why would you not want to think through your problems yourself? How could another human being be so different from me? It’s not like someone liking golf when I like soccer, because I can still relate to them on some level.
It reminds me of a study they did on domesticated dogs and wolves and I’m not an animal behavior expert but it’s interesting and stuck with me. When they were faced with a problem with human handlers present the dogs tried to figure it out for a brief amount of time and then gave up and started yelping to the humans, whereas the wolves kept trying to do it themselves. I feel like we have a dog/wolf scenario between the AI-dependent and the resistant/independent.
TL;DR get off the train girl! It’s a grid system!!
Even if you need to collate a bunch of data, don’t use an LLM for that, use a script, spreadsheet, matlab, or a compiled program that does real data analysis. Hell, even training a different NN on it would give more reliable results more efficiently.
LLMs are best when the stakes are low or for tasks where even if they give the worst possible response, it won’t affect anything.
Because there are problems out there that you’re not going to feasibly solve on your own.
Think of it this way: do you ever look up anything on the internet? If yes - why did you do that, when you could just find out the information yourself by working from first principles? Likely not only because it’s orders of magnitude faster, but because doing it yourself would probably result in significantly lower quality thank just looking it up on Wikipedia.
LLMs are like search engines on steroids. They really can help you explore problems that have already been described in the literature, so to speak.
But of course that changes how our brains need to be wired. And it’s not necessarily bad. When I was growing up, the adults were really worried that the use of search engines would degrade the ability of children to look things up in the library. And they’re right - I would indeed have a more difficult time going to the library, finding the right shelf, finding the right book, looking through the index, and scanning the pages for the information I was looking for - because that’s a skill that I haven’t practiced because I could just do a web search for the better part of my life. Doing a web search is a different skill.
That’s not to say that this is all fine and just a technological development like any other. I think we do need to be careful about how we use LLMs and when, especially in order to think through problems ourselves. I think that we were already over-reliant on the Internet before LLMs came on the scene, and it will likely get worse now. But the reason that they’re dangerous is because they are in fact really useful tools too.
That would be a good argument if not for the simple fact that ai straight up lies to you.
And its not even a question of using a “good” or “bad” model. We have observed a mathematical constant that shows no matter how large of a model you use, you can’t get below a certain percent of hallucinations.
Its far less common because models got bigger (and more Ressource intensive), but it still happens too often.
Or to phrase it in a different way, would you use a calculator if you knew every 30th answer it gives is wrong?
Very true, and adding on to what the other person said, the reason is because we are evolutionarily trained to avoid effort because it uses less energy, same reason people will watch a 30 second youtube short over reading a book.
Alongside that, there are some cases where I’ll be working on something and there’s a task I’m actually working on, and some small aside that I can’t be bothered with at that moment. For instance, I’m processing some data from a file and using it to display differing graphics depending on what’s in the file, but I don’t remember how to open it and read the data. If the llm can give me the code that does that, then I can stay focused on the graphics-creation code (and I most likely revisit the file opening later)
Destroy AI
People feel lonely and human iteration is, most of the time, horrible for many people.
I prefer to just talk to myself out loud, but I’ve seen many people resorting to AI to just chat about things, and to be honest I get the feeling of wanting something supportive and positive in your life.
Solution would be to people to stop being douchebags.
human iteration is, most of the time, horrible for many people
Not in my experience. I think you need better friends.
Your comment countering their statement literally supports it.
Is “countering” a statement horrible treatment now? Lol
Yeah, it certainly can be. It is possible to tactfully and gracefully disagree with someone without making them look or feel stupid, and without causing them to lose face. This is something our grandparents used to do for each other.
Removed by mod
I don’t understand why you need to use your grandparents as an example. In what world is a single comment on the internet the same as an abusive domestic relationship?
Some of you are replying for the sake of piling on without really contributing anything of value. This is the kind of toxicity of this site that I’m taking about.
In what world is a single comment on the internet the same as an abusive domestic relationship?
What the fuck are you talking about?
As people get older they tend to have less and less and less friends.
Mostly because we keep discarding disappointing people from our lives.
Which is better than keeping toxic people around. But it’s a shame that most people doesn’t even pass the basic “human decency test”.
most people doesn’t even pass the basic “human decency test”.
Again, not in my experience.
Ok
I don’t understand the general cynicism on this site when I see the opposite outside every day, that’s all.
There would be two logical explanations.
- Your environment is not the same as the environment of other people.
- Your expectations about people are different to other people.
That’s ok, everyone life is different.
Well, yes, but they’re also describing straight up horrible treatment. Is it really, though?
“My own personal experience is different than what you are describing, therefore you are wrong because my personal experiences are the standard.”
You do realize it’s personal experience vs personal experience, right? And that the level of mistreatment they’re describing is exaggerated, right?
I just think it’s hilarious that the toxic traits of this site are coming out just to prove their point. Maybe they’re right, thanks for confirming.
“I live in my own bubble and anything you say will be turned into a reason why I’m right”
You’re literally replying with mockery, dude.
One problem is AI interactions moving the goalposts of what folks may perceive as ‘douchebag’.
An AI providing sycophantic reinforcement at max makes someone being vaguely skeptical of you feel, by comparison, like they are being a ‘douchebag’.
Sometimes, it’s best if people in your life aren’t always acting in a way that seems supportive and positive in the moment. For example if you are having a paranoid delusion, you need a reality check rather than what feels like external validation.
Talking to yourself keeps the perspective hopefully grounded, you know who you are talking to and how full of shit they can be. The magic of AI seems to peel that grounded perspective away for some folks.
Come on man, you’ve been outside, presumably. You know that the douchebaggery out there isn’t limited to people just not being very flattering towards each other.
Of course that’s true, the issue is that when folks are surrounded by “yes men” (or AI equivalents), they lose perspective and suddenly a lack of flattery is enough to be in the “douchebag” category. Some reasonable social friction is important to keep people from going way off the rails.
So… people are born using AI, then?
If lack of flattery due to using AI is making people think everyone is mean to them and causing them to turn to AI rather than people… then…what??? I’m sorry but you’re breaking my brain here. Can you make it make sense so I can understand?
No, saying that folks who get going with AI can get into a feedback loop. Wow, this seems like a person except it likes every idea I have and never ever challenges me. Then they recalibrate to a more lower friction world.
Already happens in online communities where folks tend to get ban-happy over anything that doesn’t perfectly fit a community’s vibe. People got a bit more insufferable when they could find the most like-minded communities to endlessly reinforce their own views. AI takes it a step further by always being able to be that like-minded peer, no matter what your views are.
It’s even worse if people get hit with this a lot during their formative years, but people stay a bit malleable after growing up. Certainly some upbringings bring you closer to going all in on AI flattery than others and having a further “starting point” makes things more likely and faster, but even fairly reasonable people can go a bit nuts.
You also see it with power and wealth. I’ve known folks that at least seemed fairly grounded with a healthy relationship with being wrong or losing turn into quite unlikeable, self-centered folks after having a bit of success and excessive external validation from sycophants.
Removed by mod
