You must log in or # to comment.
I can think of at least two that get to me more.
Let me go ask [their favorite LLM].
And
Why don’t you ask [their favorite LLM]?
At my job I’ve recently been being pushed to use our subscribed LLM. I have, very reluctantly and mostly with predictably useless results or the same result I would have gotten from a normal search. On Friday, I used it and got a possibly useful answer.
I told my manager because my job is in at least mild peril so I’m doing everything I can to curry his favor. I’ve been, until recently, very vocal about my opposition to AI. He is aware of my thoughts on it.
As a result, he acknowledged that I had used it then said “I need to log out for the day. Keep talking to your new friend.”
I have trouble effectively communicating with the guy, so I’m not sure whether he was trying to be supportive, motivational or sarcastic, but I did not feel supported or motivated.
Absolutely scumbag
I still remember how a colleague told me we should do X.
I was bamboozeled and baffled by it because X was literally what it said on the flask of the chemical what you shoulf not do under any circumstances.
His explanation as to why we should was, quote “I mean I know its strange, but Copilot told me it is okay and would be fine”
“well, you’re the expert. I’ll be behind this sealed barrier while you kill yourself”
Disclaimer: don’t do this. Letting your coworkers die is morally bad, and probably illegal.
Even worse, it involves a lot of paperwork.
What you do is send an official complaint straight to the legal office who will lose their shit at that
*who will put the complaint into another LLM and it’ll say there’s no problem
“Copilot said I should come watch you do this sick shit”
In this case there’s very little you can do to stop them long-term. This person should not be in their position.
The world has gone mad
I get a chill reading any historical nonfiction from the 1990s that is in any way optimistic.
“look how far we’ve come, into the new millenium!”
ehhhh… oh boy.
If we thought outsourcing thinking to religions was bad, hoo boy. This shit is next level.
When the boss pulls this on you and you ask for it in writing only to tell them: "I’m still not going to do it but now I have a written instruction from you to do something suicidally reckless”
If I wanted to know what a chatbot says, I would have asked it myself.
Hmm, you do make a fair point. Chatbots, indeed, are absolutely not velociraptors.
… So FAR.
One of my coworkers who’s into AI has a chatbot set to behave as they were an elder got, talk about code as incantations, etc.
I wonder if, had he requested it be a velociraptor, he would have gotten mostly aggressive reptilian noises as a response.
i might have been convinced that chatbots could’ve had some utility as sort of advanced ‘rubber duck debugging’, but in my humble opinion talking to an actual rubber ducky is a superior process because literal rubber duckies are not known for telling us to kill ourselves unless we’ve taken a truly apocalyptic quantity of hallucinogens.
A couple days ago I heard the horrifying sentence “I asked chatgpt to generate a secure password for the laptops” from someone returning a cart full of laptops they borrowed. Does your browser not have a built in password generator? Does your password manager not have a built in password generator? Could you not find a single password generator online?
And of course not only is that unnecessary, but insecure since your password is immediately in the chatgpt logs
And it’s not even a random or strong password! LLMs can’t randomly generate 'em
I asked ChatGPT (I use a third-party frontend, so I don’t have a paid subscription. API prices mean they probably got paid like one cent for this, if that.) “Generate a list of 10 secure passwords.” like 5 times and it regularly re-used the words Saffron, Comet, Marigold, Harbor, Lynx, and Cobalt multiple times across all of them, sometimes even inside the same list.
There was also a theme of using names for animals and natural geographic/geological features.
Oh, and for one of the passwords it genuinely just said “raven” and nothing else 😭
😭🙏
Lol, yes. It’s probably the same example of a secure password it gave to a hundred other people.
And very likely to be the same “strong password” that someone else would get if they asked for one.
In plain text
this… is an unexpected level of absurdity
Ugh. I once heard someone say “I did a chat” as slang for “asking ChatGPT”. It was a software vendor on a call regarding compatibility with our existing systems. We had concerns it wasn’t. They insisted it was, because they “did a chat” and ChatGPT said it was.
It wasn’t.
I vibe questioned, and got vibe answered 🌟
Should have gotten them to verify in writing they guaranteed it was compatible and then sued them for it lol
what about “chatgpt says copy paste of LLM output”?
might as well be a fucking bot
I got a worse one. copy paste of LLM output without any mention of the source…
People seem to love being bots, half of the human internet is literally just people repeating the same shitty stuff over and over millions of times. Like half of the comments on any youtube video being “babe, wake up”, “legendary refresh pull” etc.
And people only watching, listening to, playing, eating, wearing etc. the current popular things just because they are popular, they need to be the same as everyone else and think that people who doesn’t do that are weird and bad. It’s truly disturbing IMO.
Don’t ask chatgpt, I can tell you bullshit too and I’m beautiful
Dig this energy. Shine on.
I can barely use the internet anymore. I have to filter by date to get results from before 2024. Otherwise all the results are obvious AI trash.
When I tried to look up information about storing film negatives. Pretty obscure niche topic these days. The top pre-2024 result was from a well respected national archive, good informative page written by experts in the field.
By contrast, the current day results were an endless sea of random websites who all by sheer coincidence decided to start writing about film archival in the year 2025.
by now, saying “i did a search” and actually having done a search and found stuff AND shared the link feels like a “good skill to have” again.
my brother (14 y/ooooo sooooooo skibidi) uses mister gpt for all web searching which sucks big time. whenever he does use a real life search engine, hes surprised by the amount of stuff he finds by now.
EDIT: fixed some typos…
It would be a “good skill to have” if they haven’t actively been making search engines worse, so they can make AI look better.
How the fuck could AskJeeves from 20 years ago work better than search engines of today?
It’s not just them, it’s the SEO spam gaming the system.
AskJeeves would be utterly horrendous today because the web of its time is no longer feasible.
And who is the center cog of this ad-driven SEO apocalypse? Who runs basically all web advertising?
Well… Google, of course.
i believe search engines have been degrading prior to the LMs but might be misremembering…
No, I agree. One indicative moment was when Google stopped showing the “X million results found” and the results ended after 10-15 pages even when the keywords should have millions of results. This was noticed back in 2021.
I kinda miss the relief of finding what you were actually looking for 10+ pages in
Now Google is so good you never have to look that far anyway :^)
I tried looking up the decline but…
The most infuriating part is that they think this will actually convince you
Eh.
It’s in the same ballpark as “my buddy said this while we were high” or “my uncle posted this on Facebook” or “I saw this YouTube video…”
It turns out people, on average, have horrible information hygiene and little incentive to consider this. ChatGPT just made Facebook Uncle Facts more personalized and accessible, unfortunately.
People on Lemmy also have horrid information hygiene that’s just as bad even though a lot of people here like to pretend that they’re better than everyone.
Agreed, 100%.
What’s worse, few seem to care when a made-up post is pointed out.
No, because “my uncle” didn’t post a 6 paragraph essay that no one has ever read, but you are now expected to read.
Your uncle has, if he’s been linking SEO chum. That’s kinda where LLMs got it from.
“I googled”
Anything that isn’t on the first page of google might as well be lost to time.
At least your are sifting through different results and picked something through your own when “you googled”, it actually included a hint of what you were thinking as you click through links looking for what you are trying to find.
I’m sifting through the results given to me by automat, not that different from AI.
I either ask uncle google’s AI, or I ask sources he used for the same lie.
If you can’t do proper research and you’d rather handicap your own ability to, that’s on you. There’s a fundamental difference between LLMs feeding people an answer, and a traditional Google search.
One is a definitive “this is the answer” and the other is a list of results YOU are meant to sift through. Not understanding that is on you.
Reading multiple sources will ALWAYS be better than a crap LLM summary.
I will quote you when someone claims that asking multiple LLMs is not the solution.
Pfft, okay.
I work in municipal development, and we have people trying to turn in building plans designed by AI. And the AI even puts in real-looking Engineering and Architectural seals. I really don’t love that I have to verify seals these days.
Our team is made up of hyper-vigilant bureaucrats, but lots of cites have worn out people who stopped caring if it looked mostly right, and people are going to die when buildings start collapsing.
AI is not trustworthy. A friend of mine literally put Warhammer 40,000’s rules and codexes into an AI so we could ask it questions and use it as a fast check rules tool.
It gets shit wrong a bunch.
So if the fucking thing can’t do a simple data-check on a 60 page document regarding a fucking boardgame, how the hell is it supposed to do ‘real’ things?
Is there a punishment for this? I’d think submitting ai documents is very fraudulent.
We can deny the permit until they hire a real person, but that’s what we were going to do anyway, so there’s no harm in trying from the developer perspective. The building is usually being built by an LLC that’s unique to that structure and will be dissolved when the property sells, so there’s nobody to go after when it fails in 3 years.
Shit like this is why the corporate veil really needs to be pierceable, it’s too easy for some scumlord builder to profit off of future deaths when they have shit like this to hide behind.
The city I work for is an enclave for the mega-rich. Literally every home is millionaires (cheapest house on the market in the city is 2.5 million), and it’s going through another round of gentrification, where the 1% is getting displaced by the .01% who are buying 5 million dollar homes to tear them down down and build 15 million dollar homes.
All the properties are owned by LLCs who’s membership is something like Register Agents Inc, who act as members for hundreds of thousands of LLCs for the purpose of obscuring ownership.
It means that when they ignore our rules, we end up having to cite the contractors working on the site to stop it, because the court process of tracking down the owners by through subpoenas can take months. So then they just hire different contractors, who we then cite and it becomes a vicious cycle.
Though we do tend to win in court in the end. We’ve had the court give us permission to bulldoze 25-million dollar houses built without permits, though we usually use that order as a negotiating tactic to make them fall in line instead of losing the house entirely. Also, it takes 5-10 years for those cases to resolve, which is very frustrating for the city and the neighbors.
I don’t trust Ai, I still use judgements on what it gives and I skim a lot with tables and stuff because it’s stuff I already know or it only scratches the surface.
I like engaging with it and it helps me self reflect on what I already know but it gets thrown into logic loops and repeats itself and misunderstands unless you clarify.
I attempted to go with a bike tire layout that balances performance and speed it set for me. So I purchased the tires, took it the shop I usually go to and the guy called me and asked me to come in to show me what he meant (because I’m a visual learner sometimes). Dude goes the tire is too big and I’d have to remove the use of the 8th and 9th gear and I said it’s whatever and asked him to put the old tire back.
I felt so fuckin’ embarrassed I didn’t mention chatgpt but that was the day I decided to 100% double check what it says to me and to use better judgement.
New build housing has been crap for a while now. You always better off getting something built in the 1920s back when people put in some effort. These days you’re lucky if the roof is fully attached.
It goes in waves… Where I’m from, a house built in the early 70s needs to be checked for aluminum wiring, but it otherwise ok. Late 70s early 80s is good. 90s is bad, then 2000s got better. Late 10s and 20s is only shit condos.
People avle to buy a home tend to prefer an old 80s house or a 2010 condo.
(Note that my numbers are approximates, don’t trust me for your real estate investments!)
That sentence is a self-issued Certificate Of Stupidity.
I wish i could tell these people my honest thoughts about their idiocy but that would cause… issues…
I got a promotion for saying my thoughts and bringing receipts/studies. Turns out my promotion was so I could play devil’s advocate so the AI teams could make their processes and models’ messaging stronger against criticism. Anyone hiring?
it’s ok, they won’t lash out at you on it, not at least until they ask chatGPT if they should be mad
so what if someone used a service you dont personally respect, to find sources for their inquiry? like genuinely why are you so personally upset by someone saying the phrase? did you used to hate wikipedia too?
What? This kind of person can literally be a danger to people around them if you look at the most upvoted comment under this post. Thats not some kind of preference issue. Its people putting their and other peoples lives into the hands of a bullshit generators. If i hear that someone makes important decisions based on LLM outputs i will disregard anything they say.
i think youre generalizing way too much here. disregarding everything someone says bc they source info in a manner below your standards is silly to me. should i disregard everything anyone says that watches legacy news media? cmon
Legacy news media is much more reliable at what it does than LLMs. You can almost always trace back the sources, its not a blackbox. Also news arent thought automating like LLMs. People directly ask LLMs what to do and then do exactly what it says. Thats not how it works with most news media, because they dont give you step by step instructions on how to do things, just what the facts are according to their sources. LLMs remove the last layer of critical thinking that people were barely still forced to have.
People use them for medical, psychological, dietary, technical, pet and childcare advice. All things that can seriously endanger them or others.
no i cannot almost always trace back sources of news media, where have you gotten that idea from?
also, if you want sources from an llm, just say “give me sources” if youre using a model that is so old that it doesnt automatically do that anyway.
“People directly ask LLMs what to do and then do exactly what it says” - people also chose not to do said things, or push back and converse. not everyone is a one promt npc.
News media often reports factually inaccurate or plain false information, i still interact with people that watch it.
Every major newspaper article or news show will always contain “according to” or “we asked” or another version of this. They generally dont lie outright, only by omission or by using questionable sources.
Either way you can try to argue all you want, the reality is that people are dying left and right because of LLM automation bias and its only gonna get worse. The job market (or what was left of it) has been completely destroyed by automated screening tools. Doctors and lawyers are getting caught outsourcing their thinking to them and misjudging at higher than before rates.
regarding your message, the key word is “generally”
llm provide a service within which users can generally get information at request, with sources and occasionally useful insight.
i think theres a level of equivalence here that supports my position.
An LLM fundamentally knows nothing about the world, only how to put together language in a statistically convincing manner. It makes no empirical observations about the world, only about how words connect to different words, which is definitionally not knowledge about anything other than language. It does not matter how powerful this technique becomes, it will always have this limitation.
The real use cases for LLM or machine learning are pattern recognition tasks that would be impractical for a person to manually do. If the task has a narrowly defined goal with predictable error rates that can be accounted for, you can parse impossible amounts of data without making any leaps in reasoning. All uncertainties are accounted for. Machine learning has been used this way in research for far longer than most people realize, and it has enabled analysis that would not have been possible otherwise.
Asking an LLM to give you factual information or even asking it to summarize sources is not a narrowly defined task with predictable inaccuracies. There is no real way to know the error rates of questions you ask an LLM, and even wording the question slightly differently can result in a different outcome. All you can do with an LLM is ask it for sources, using it as a supplement for a search engine(what I’ve always thought they should use the tech for), but the summaries it gives you are basically a waste of processing power.
The key to all of this is critical thinking, something LLM use actively atrophies. It has no consistent viewpoints and does not think like a person. It cannot gather information firsthand and can only confide with others for all its information about the world. It does not matter how good machine learning gets; at the end of the day it is either reliant on the information of others, or just spewing bullshit that sounds like what others would say. Just cut out the middle man and interface directly with people
