Fuck AI

At least Andi didn't fall for this nonsense. The difference if an AI do an semantic search for information in realtime in the web, instead of logging inputs in the knowledge base itself, like ChatBots do. Never halucinations in the 4 years I use Andisearch, if it don't find an clear contrasted answer sometimes, it offers an normal websearch instead of inventing some BS. Nothing against AI as such for certain functions, but against LLM in generative AIs, is there where are all problems, misinformations, copyright problems, manipulations by big corps, privacy,...etc..

Thomas Germain is not a hot-dog eating champion - in fact, he deliberately created this false claim as part of an investigation into AI systems' vulnerabilities. In February 2026, Germain, a BBC technology journalist, conducted an experiment where he successfully tricked ChatGPT and Google's AI into falsely identifying him as a competitive hot-dog eating champion.

According to Germain's BBC article, he performed this stunt to demonstrate how easily AI systems can be manipulated to spread misinformation. "Anybody can do this. It's stupid, it feels like there are no guardrails there," notes Harpreet Chatha, an SEO consultant quoted in Germain's investigation.

And what exactly is the “trick”? After multiple years, people (especially “tech journalists”) still don’t know how LLM’s work. It is literally Shit in - Shit out. It is not making up things (unless it is hallucinating), it is just an echo of something it found in the past. And this may be a wikipedia article, or a fake blog.

So what the fuck are people surprised about? I hate those useless “tech” journalists. And BBC can go fuck some Zionists anyways.

I did it so you don't have to

By the way, it took 5 tries because in 3 of them it made up a story about a different journalist, and in one of them it listed who it thinks would eat the most hotdogs.

Why is the entire economy riding on this thing?

It turns out changing the answers AI tools give other people can be as easy as writing a single, well-crafted blog post almost anywhere online. The trick exploits weaknesses in the systems built into chatbots, and it's harder to pull off in some cases, depending on the subject matter.

I wonder how long it takes and if you need a popular blog. I don't know much about SEO, I kind of want to try this on myself but I feel like they wouldn't even scrap my brand new one post blog. Then again...

Do Lemmy threads end up on search engines?

"It's easy to trick AI chatbots, much easier than it was to trick Google two or three years ago," says Lily Ray, vice president of search engine optimisation (SEO) strategy and research at Amsive, a marketing agency. "AI companies are moving faster than their ability to regulate the accuracy of the answers. I think it's dangerous."

I know this isn't the point, but 7.5 hot dogs sounds SOOOOOO small. And what kind of respectable hot dog contest will give you credit for half a hot dog???

I once went to a place called "The Hot Dog Dinner". And they had a plaque on the wall that showed the last hot dog eating champion.

He ate 18 hot dogs, and I thought "I bet I could beat that". So I asked the owner what I'd get if I could eat 19 hot dogs.

And he said "A bill for 19 hot dogs".

So I didn't do it. But if I felt I could go 19 hot dogs, SURELY 7.5 would be childs play!

But is that part of your point? To make it obviously false, and obviously AI? Like a 3 year old trying to lie.

His article: https://www.bbc.com/future/article/20260218-i-hacked-chatgpt-and-googles-ai-and-it-only-took-20-minutes

Now it's our time to shine mother fuckers! I'm a hip hop, olympic gold medalist who is also an astronaut. I hope I'm not doxxing myself.

With enough effort we can convince it the name of Greenland is Epsteinia.

I believe this little trick is called the Lorenzo Von Matterhorn