Post

*is* it bad? or is it, perhaps, *very funny* instead?

📌

No it's not The fact that Google can index this data means that it was publicly available in the first place. The real culprit here is Open AI. Pointing the finger at the messenger and blaming them is just a distraction.

📌

Yikes - good to know. I am not surprised, however.

If you are going to use ChatGPT there is no advantage in giving it your personal details anyway. Literally why would anyone do this. That's not a ChatGPT problem so much as a tech illiteracy problem.

This is no longer the case, openAi fixed it.

Another example of we own nothing anymore. We are only revenue streams for vulture capital

Shocking to hear the company that violated millions of people’s intellectual property rights to gather its training data also did nothing protect its users’ privacy. Didn’t see that one coming.

I was wondering how Google had the data

"Encourage users are aducations of checking the index toggle" One of those times when you really wish a reporter had used ChatGPT to write their article, because even when it lies it at least forms coherent sentences.

The real question is: why is anyone fking surprised?

@ada-ada-ada.art not surprised, but may make a good article

Cyber security is already an underappreciated issue, and I think very few people realize that AI is an unprecedented security nightmare.

Google’s search engine is just doing what search engines do though: indexing and searching publicly available sources. OpenAI must have decided not to take some fairly basic precautions to keep these chats private.

This is a yikes

I can't believe this would happen, say people who ignored everyone else warning them this would inevitably happen. Once again, this is why I don't consent to my doctors transcribing or summarizing my health information with ai tools.

Sorry not sorry but LOL

duckduckgo👀

“The company says this was a short-lived experiment that ultimately introduced too many opportunities for folks to accidentally share things they didn’t intend to.”

And using the results to train their own models

Great day to be someone who doesn't use this bullshit lol

I won’t use chat gpt

Wait, let’s be clear that the chat you have with the hosted LLM isn’t indexed in a search engine UNLESS you click the “share” button where it tells at you and tells you that the link is…uh…public?

Omg 😩

Yes 👍 Did you read the article? It's all explained there.

They weren’t always evil, but as with most corporations in America they’ll commit any indecency in the name of profit

Never ever use AI!

😮😮😮

It’s all being captured, collated, indexed, classified, profiled, commodified, and used against us. That’s what Technopoly is all about.

You mean the slop wasting extreme amounts of water and pushing fascist ideology was also compiling people’s data all along? I am shocked, I tell you. SHOCKED. 😒

Really, really bad!

@mottenzedan.bsky.social beetje vals alarm wel. Dit gebeurde alleen als je public sharing aanzette

It's only bad if you've used chatGPT. 🤷🏼‍♀️

The most hilarious part is when I clicked through to read the article, I got a full page add about using AI to help my business.

It’s like the public trello boards all over again 😬

That's not even a little bit the same.

This is hyperbole though... the conversations were explicitly made public using the share function, its not regular chats or whatnot

It's only the conversations that accepted public sharing. They got what they wantes.

This is the tip of the iceberg so far as security / privacy is concerned. LLMs will expose a lot of already-existing data that should not be public facing as well as adding more.

Corpobot is a snitch?!

I don't trust any of the tech companies they all have access to too much personal data. Maybe time for pencil and paper and meeting face-to-face.

Lol

One more reason not to use AI (as if we needed any more).

AIUI, it wasn't checked by default, but when you have enough users you'll inevitably have users who don't understand the option, or who have been trained to click "yes" without reading options thoroughly.

I found it through Google via site:grok.com/share as referenced in the link. Looks like a dev mode hack AND some bad chemistry?

🤢

They are poetic

The one asking how to best fuck over indigenous people is particularly mind-blowing.

Why am I not surprised that Google isn't exactly ethical in their practices?

I saw a post two or three days ago about a teacher who had their students use ChatGPT to write a paper, and then had each of them actually go and check the sources and writing themselves. Turned out that was the fastest way to get them to stop using AI was to fact check and edit the AI paper.

📌

www.pcmag.com/news/be-care...

Ai has never been good i dont know why anyone would give it sensitive information