Earlier this summer, rangers with Shasta-Trinity National Forest found a woman camping in an off-limits area on the south side of Shasta clearly marked with “no camping” signage, Meyers said. She told rangers she had asked ChatGPT where she could camp in the forest.
“AI told me I can camp here,” she said, according to Meyers.
Then in August, while hiking in Shasta’s popular Bunny Flat area, a man stopped rangers to ask for directions. He showed them a print-out map he’d been following and explained that he’d prompted ChatGPT to produce it for him.
“It was terrible — terrible — not even close to accurate,” Meyers said. “It had route lines and place names that were totally off — just ridiculous. We were like, dude, you gotta throw that away.”
From the latest mishap:
During the hike down, the young climbers revealed to rangers that they’d relied on ChatGPT to guide their planning. Authorities noted that as a “key factor” of their mishap.
The climbers “reported that AI recommended simple carbohydrates for their climb, rather than fats because they ‘take too long to digest,’” according to the official report. “In their words, ‘We relied too much on AI rather than our own critical thinking.’”
Edit - from the non-Archive version of the article:
Correction: A previous version of the story misstated the AI tool the climbers used to plan their trip. It was Google Gemini.
I wanna see a slow burn techno horror movie where this is how the “AI” starts killing off humanity.
“It started with ‘gluing shredded cheese’ to pizza… and now its telling us to bottle feed babies arsenic!”
It doesn’t do the “Skynet makes killer robots thing” or “launching nukes thing” it just tells us with total confidence to poison ourselves or drive into a lake or camp on a frozen mountaintop… and a huge chunk of humanity just … doing it.
This isn’t anything new, is just techbros at the reigns now instead of the church or nation state.
How do these people actually manage to see the map an AI generates and not be suspicious immediately? I’ve sent the shit it does, did they not even once wonder why the fucking path went a different direction or why the fucking heights were different than on the map? No offline GPS? No offline digital map? Google maps nowadays caches wherever you are, how badly do you have to go, how much do you have to trust AI, to go and essentially fall off a cliff at that robots discretion?
This shit is showing around some people I know, people I work with. I want nothing to do with them because holy shit they will eat belladonnas.
because holy shit they will eat belladonnas.
That’s probably the easiest way that a poisoned dataset could kill someone.
“It was terrible — terrible — not even close to accurate,” Meyers said. “It had route lines and place names that were totally off — just ridiculous. We were like, dude, you gotta throw that away.”
TAKE IT AWAY FROM THEM!
“You don’t need long lasting food if you’re going away from your food cache” why the fuck would you believe the AI if this is the summary of what it told you? Because that sounds like the exact opposite of how humans have operated since the Stone Age. “Yeah take these simple carbohydrates you’ll burn through in an hour, when you’re out away from replenishing your gut” like what the fuck
Edit: it’s unsurvival advice
Confiscate the phones of these goobers
Anyone who has ever had a physical job knows how much of a difference it makes to get some slower burning energy sources into your diet (I liked an avocado at breakfast to keep me going)
Two things should happen every time something like this happens. First off, the people who did it should have to pay in full whatever costs their actions carried. Next, the company who owns and runs the AI that produced the result should be required to pay 10x whatever the costs.
At some point it stops becoming a cost of doing business and starts becoming a real problem for the AI companies. Until that point they literally will not give one single fuck. Punish the dumbasses who use AI and punish the companies proportionally too. Make it hurt both sides of the equation and it’ll definitely stop happening
I’m all for penalizing the business but we do not want to be charging people calling for help unless they were actively malicious. It’s already a problem that people fail to call an ambulance because they’re worried about the cost. If someone has an accident while hiking we don’t want them doing the calculus of, was I too much of a dumbass in planning this? Human error is very often a factor in rescue situations and we don’t want people whose lives are at risk worrying that calling for help might demolish their family finances.
How about a permanent ban on use of AI.
You fuck up this bad, you get AI Paroled for the remainder of your mortal existence and Google isn’t allowed to show you AI results if you’re logged in.
AI Paroled for the remainder of your mortal existence and Google isn’t allowed to show you AI results if you’re logged in.
Don’t tempt me with a good time
I think legally forcing companies to make it possible is a great way to wedge things open and make it an available option for everyone.
“If you could do it for that guy, why can’t you do it for me?”
“Why do you make this so difficult to disable?”
“Make it a user setting.”
“Disable it by default.”
“Get rid of it entirely.”
not harsh enough, needs to be some kind of forced amish situation where you’re not allowed to use anything more advanced than a graphing calculator. give 'em the mitnick
Oh definitely
Disappointed it didn’t include a picture of the terrible map, that would have scratched a certain slop itch
me no read, robot read. thank you robot.
there are some people whose brains will be useless except for writing prompts.
Will they learn something from this?
Probably not. If these three dumbasses have already outsourced this much thinking, they’re likely never getting better.
No, you’re right that they won’t - it’s in their response i.e. “We relied on AI ‘too much’”
Authorities noted that as a “load bearing seam” of their mishap.
smdh; didn’t even petition psychic bigfoot for aid
The whole situation with LLMs is so incredibly infuriating. The technology itself is genuinely cool. But the fucking greedy tech bros make it actively worse by training it to pretend it was some autonomous artificial intelligence that it so clearly is not. And due to all the consequences from capitalism milking the shit out of it the whole fucking society goes to shit and now you are left with basically two factions: on the one hand you have the evangelicals that honestly think that LLMs are the best thing since sliced bread and it will revolutionize the world and on the other hand you have the haters that think LLMs are a total scam and just as bad as any other autocomplete, which has been around for decades.
The truth is somewhere in the middle. The tech is actually pretty impressive but the tech companies are overselling it hard and train the LLMs to oversell themselves hard to impress investors and the public. This is the real reason for all the AI psychosis shit. You cannot blame gullible people for being gullible when the tech is actively designed to trick people. Don’t blame the people, don’t blame the tech, blame the fucking assholes that are responsible: The greedy fucking tech bros.
I personally use LLMs pretty much every day and i have personally also used it to plan trips. But i do know that it makes shit up all the time and i have to sift through the annoying sycophancy and useless rambling to get to the actual useful bits: A very efficient way to search for information that i know is probably widely available in books and the internet but would take me much more time to find manually.
I seriously think that this technology could be so much more useful if it was optimized for those kinds of tasks and would not pretend to be a sentient artificial intelligence all the time.
Future Darwin award winner
