Grok behauptet, dass die Sicherheitsvorkehrungen verschärft wurden, nachdem Benutzer KI missbraucht haben, um Bilder von Frauen und Kindern zu verändern
https://www.cnbctv18.com/technology/grok-claims-safeguards-tightened-after-users-misuse-ai-to-morph-images-of-women-children-ws-l-19811512.htm
16 Kommentare
„xAI has implemented strict guidelines to prevent Grok from generating explicit or non-consensual content,“ **the tool wrote in an automated response to a post by CNBC-TV18.**
Telling that xAI is making Grok saying it rather than a real person
It’s almost like these safeguards should be built into these tools *before* they’re made available to the public. But what do I know? I’m not a billionaire.
Rapidly coming to the conclusion that „AI“ adds nothing of value to society and is instead extremely detrimental in a huge variety of ways.
Previously grok was telling children to send nudes. Seems to me leadership of X and grok should be under investigation
Oh look, yet another predictable thing happened with no attempt whatsoever at prevention.
I, for one, am surprised that a pedophile billionaire created an AI chatbot that creates CSAM.
If Musk makes money on it, expect no safeguards.
Something something free speech blather blather.
Worst „AI“ ever
Why the fuck would I believe Grok?
Crok isn’t a person. it’s a damn computer algorithm. it doesn’t „Say“ anything. it’s a damn line of code. someone programmed it to say that. stop attributing something Xtwitter does to a brain dead computer. and stop trusting anything elon musk and Xtwitter does or says.
>Grok claims
Into the trash it goes.
It’s literally insane (if you go by Einsteins definition of insanity and repeating the same things over and over) that AI companies can get away with aiding toxic and or criminal behaviour over and over again with no adequate punishment. If this were any other technology (P2P) or person they’d be charged, tried and punished in prison by now.
Anthropic has failed to successfully introduce safeguards to Claude that would prevent it from letting people burn more money than their subscriptions are worth. They’ve tried, but every attempt has failed. If these companies can’t limit how much money their models are burning why would anyone believe they can limit anything else they do?
It’s almost like pedophiles are way more widespread and common than anyone wants to admit. That needs to be addressed at a deeper level than what we currently do which is ignore it or say kill them all.
AI doesn’t claim anything, it just reposts.