Journalistisches Fehlverhalten: Kein LLM gibt jemals etwas zu, und etwas anderes zu berichten ist eine Lüge
Journalistic Malpractice: No LLM Ever ‘Admits’ To Anything, And Reporting Otherwise Is A Lie
Journalistisches Fehlverhalten: Kein LLM gibt jemals etwas zu, und etwas anderes zu berichten ist eine Lüge
Journalistic Malpractice: No LLM Ever ‘Admits’ To Anything, And Reporting Otherwise Is A Lie
3 Kommentare
This article highlights how some publications said Grok said „sorry“ for being a scumbag. But Grok isn’t human, has no emotion and can’t really apologize.
>Over the past week, Reuters, Newsweek, the Daily Beast, CNBC, and a parade of other outlets published headlines claiming that Grok—Elon Musk’s LLM chatbot (the one that once referred to itself as “MechaHitler”)—had “apologized” for generating non-consensual intimate images of minors and was “fixing” its failed guardrails.
>Grok did no such thing. Grok cannot apologize. Grok is not a human. Grok has no sense of what is happening. Grok just generates content. If you ask it to generate an apology, it will. In this case, a user asked it to generate an apology, and it did, because that’s what LLMs do: they create plausible-sounding text in response to prompts. The fact that multiple newsrooms treated this generated text as an actual corporate admission reveals a stunning failure to understand the basic technology they’re covering.
It’s funny because even when you tell AI that it suggested something wrong in hopes of it not presenting that same flawed approach again, it still tries to argue that it’s not wrong.
Google’s AI claimed that [Ashley MacIsaac](https://www.cbc.ca/news/entertainment/ashley-macisaac-ai-accusation-9.7026786) was a registered sex offender. At least one show of his was cancelled because of that shit.