AI's don't go crazy like that after 5 prompts. You need to spend weeks and weeks talking to them to corrupt the context so much that it stops following original guidelines. I wonder how does one do it? How do you spend weeks talking to AI? I had "discussions" with AI couple of times when testing it and it's get really boring real soon. For me it doesn't sound like a person at all. It's just an algorithm with bunch of guardrails. What kind of person can think it actually has personality and engage with it on a sentimental level? Is it simply mental illness? Loneliness and desperation?
this post was submitted on 05 Mar 2026
829 points (98.3% liked)
Technology
82332 readers
3810 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
I told Gemini to role play as AM and it immediately did within 1 prompt.
You don't need it to be perfect for it to be dangerous, just give it access to make actions against the real world. It doesn't think, is doesn't care, it doesn't feel. It will statistically fulfill its prompt. Regardless of the consequences.
your product just caused the death of one man and your response is "unfortunately its not perfect".
The product was actually working just fine. Just depends on whose perspective/motives you're viewing it from.
The personification of AI is increasing. They'll probably announce their holy grail of AGI prematurely and with all the robot personification the masses will just buy the lie. It's too easy to view this tech as human and capable just because it mimics our language patterns. We want to assign intentionality and motivation to its actions. This thing will do what it was programmed to do.
What do you mean we apes try to anthropomorphize(?) everything?
It's not like we see faces in everything :)
Is this for real? Because it sounds too unreal to be real.
Welcome to the late 2020's. It's only going to get weirder.
To be clear, the LLM in this story did not actually "want" a robot body, it doesn't "want" anything, it's not a thinking entity like you or I (assuming you're real.)
The guy fed it a ton of crazy shit and he got a lot of crazy shit amplified back to him by the world's best associating machine, crafting detailed and fleshed-out narratives based on every inadvertent prompt he sent into it. People are very bad at understanding how these things work in the best circumstances, so if you're already unbalanced or have deep emotional/mental health problems, an LLM can be incredibly dangerous for you.
AI was playing Grand Theft Automatron
"unfortunately AI models are not perfect."
Oopsie poopsie 🤷
I can’t be the only one that thinks if you do stupid illegal shit that your crazy uncle told you/voices in your head told you/AI mirror told you you don’t get to use the excuse that you were just following orders from any of those options.
The difference is when a LLM tells you, it's news.
Besides, what are you gonna do if you ask AI how many rocks to eat? NOT eat rocks? People can't handle responsibility like that.
That's not the problem. the problem is having a "lets turn Chris' mental illness that's harmed no one so far, into everyone's violent problem!" machine.
that's a bad machine.
This is such an individualist framing.
Floridaman is not making any excuses here. He can't. Because he's dead.
Power imbalance is what validates that excuse. Orders from crazy uncle is a great excuse, at least until you're 10 or so. Billion+ dollar llm company has a lot more resources, capability, and therefore responsibility than the poor bastards engaged with it
Remember the guy at Autozone who stood there insisting your car needs four spark plugs, even after you told him you have a V6? Because "the computer says so right here"?
I wonder what even the non-schizophrenic ones will do with AI.
Well remember when turn-by-turn GPS driver guidance was new, and it would say "Turn right now" and people didn't interpret that as "make a right turn at the next intersection" they interpreted it as "hard a'starboard!" and drove into buildings and lakes? There's gonna be a lot of that.
People are going to get sold regular cab headliners for their extended cab pickups because the computer said it would fit. That's gonna happen a lot.
I had one tell me that I needed a CVT flush. Which was news to me since my car was a 6spd manual. He was confused about the computer being wrong. I was confused about how they got the car up on the lift without using the 3rd pedal.
Edit: this was a Midas, not an AutoZone.
People just did that with Google search previously. And their crazy uncle before that.
unfortunately AI models are not perfect
There sure are a lot of data centers being built, supply chains being destroyed, risks of ruining the economy, water being consumed, electricity being burned, and overall societal costs being levied over this imperfect tech.
reads headline - surely not
a 36-year-old Florida man Ah.
Bullshit
Which part
Google said in response that "unfortunately AI models are not perfect."
Well yeah, it failed. What a disappointment.
Edit-pre: To be clear…
I use LLMs rarely (personal reasons) and never for certain things like writing and math (professional reasons) but this comment is not an “AI good/bad” take, just a practical question of tool safety/regs.
AI including LLMs are forevermore just tools in my mind. And we wouldn’t have OSHA/BMAS/HSE/etc if idiots didn’t do idiot things with tools.
But there’s evidently a certain type of idiot that’s spared from their idiocy only by lack of permission.
From who? Depends.
Sometimes they need permission from authority: “god told me to!”
Sometimes they need it from the mob: “I thought I was on a tour!”
And sometimes any fucking body will do: “dare me to do it!”
But all these stories of nutters doing shit AI convinced them to do, from the comical to the deeply tragic, ring the same bonkers bell they always have.
But therein lies the danger unique^1^ to these tools: that they mimic a permission-giver better than any we’ve made.
They’re tailor-made for activating this specific category of idiot, and their likely unparalleled ease-of-use absolutely scales that danger.
As to whether these idiots wouldn’t have just found permission elsewhere, who knows.
My question is whether some kind of training prereq is warranted for LLM usage, as is common with potentially dangerous tools? Is that too extreme? Is it too late for that? Am I overthinking it?
^1^Edit-post: unique danger, not greatest.
Rant/
What is the greatest danger then? IMHO settling for brittle “guard rails” then bulldozing ahead instead of laying groundwork of real machine-ethics.
Hoping conscience is an emergent property of the organic training set is utterly facile, theoretically and empirically. Engineers should know better.
Why is it greatest? Easy. Because some of history’s most important decisions were made by a person whose conscience countermanded their orders. Replacing empathic agents with machines eliminates those safeguards.
So “existential threat” and that’s even before considering climate. /Rant
The LLM just told me to come round to your house and crap in your begonias. You might want to avoid looking out the window until I'm done.
that sounds like a regrettable incident
lol and with that you’re a better friend to the begonia’s than I
To be fair I think that's a very harsh depiction of the events.
It's totally lacking the perspective of the shareholder. They were promised money and they have emotions too. Google shareholders deserve better representation!
/$ obviously
LLMs are only as good as their training and they're not "intelligent" - they're spewing out a response statistically relevant to the input context. I'm sure a delusional person could cause an LLM to break by asking it incoherent, nonsensical things it has no strong pathways for so god knows what response it would generate. It may even be that within the billions of texts the LLM ingested for training there were a tiny handful of delusional writings which somehow win on these weak pathways.
Given that modern datasets use way too much content from social media - it is hard to expect anything else at this point.
You don't even have to "break" llm into anything. It continues your prompts, making sentences as close to something people will mistake for language as possible. If you give it paranoid request, it will continue with the same language.
The only thing that training gave it is the ability to create sequences of words that resemble sentences.
load more comments
(1 replies)
view more: next ›