ChatGPT users shocked to learn their chats were in Google search results
OpenAI scrambles to remove personal ChatGPT conversations from Google results.
(Me: nothing to add here, it'd be like dynamiting fish in a barrel.)
Discussion
Loading...
@cstross Some people are shocked to learn that employers are restricting access to ChatGPT and other LLMs. Some employees are using personal phones as mobile hotspots to get around that, but will probably be told "No, we *really* mean no access!"
On-site instances of LLMs are allowed, but I expect the cost of those to head towards the stratosphere soon.
@cstross Use Lumo or Le Chat
@cstross hahaha
@cstross
So to be clear: people are shocked, that when they ticked a box saying "Make this chat discoverable" and the description that the chat can then appear in web search results, that these chats are actually discoverable by a Google web search.
So to be clear: people are shocked, that when they ticked a box saying "Make this chat discoverable" and the description that the chat can then appear in web search results, that these chats are actually discoverable by a Google web search.
@cstross What us troubling is that these big IT comoanies always put their ibterests way ahead of tgise who use their services. It is immoral, unethical, and damaging. They operate in the "wild west".
@cstross I just won the betting pool at work.
Might also be fired for it.
@cstross so, than i shouldnt use it again, thanks.
@cstross I'm getting flashbacks to when people were shocked their Tweets were indexed 💗
@cstross I wonder why everybody acts surprised.
Wasn't that forseeable?
Wasn't that forseeable?
In the quest for data the big companies don't stop at anything to get ahold of it. They even break laws to do so.
It seems that OpenAI ruddered back and disabled the ability to search chats via Google
So, still don't see the pattern? They always try some shit first and exploit it until someone complains. Then they draw back and act surprised how that could have happened.
@cstross No comment …
@cstross haha !
@cstross
Euwwwww.
@cstross Is anyone surprised? Honestly not surprising at all.
@cstross
What evidence is there?
What evidence is there?
@cstross Open AI are aware that they have to earn back trust....
They continue to steal resources from every server they "index". They stole all of humanities digitised content but they can "earn back trust".
@cstross
"I'm shocked, _shocked_, to find Google going on in here."
@cstross
*pretends to be shocked noises
@cstross More like:
Users shocked that clicking a box labeled 'Make this chat discoverable. This allows it to be shown in web searches' makes chat discoverable, and shown in web searches.
🤦
@cstross It sounds really easy to shock a chatGPT user 🤣
@cstross
“According to Fast Company, that included chats about their drug use, sex lives, mental health, and traumatic experience”
What. The. Actual. Fuck. is wrong with people that they’re “talking” to a chatbot about stuff like this?
@mark_f_lynch A lot of people fall for the ELIZA delusion—if something can talk, then it's a person. This is a very human cognitive failure mode, being exploited by depraved capitalist grifters.
@cstross @mark_f_lynch they reveal those things because it's not a human
@cstross
There is no "removing" from a LLM / neural network.
There is no "removing" from a LLM / neural network.
You‘ll have to re-train everything again - at least from the time just before you added the part in question.
Or you keep the mistrained network and add output filters.
@cstross So are they saying Technology Minister Peter Kyle's inane conversations with his robot friend are just out there for anyone to find?
@cstross
wait till we see the shock on their faces when they learn that not everything on the internet is true.
wait till we see the shock on their faces when they learn that not everything on the internet is true.
@cstross
What’s the fuss? It works as designed. Everybody knows that.
bonfire.cafe
A space for Bonfire maintainers and contributors to communicate
Automatic federation enabled