Grok refuses to answer a prompt, says its a violation of the "OpenAI Policy"

Comments

• brsrklf@jlai.lu ⁨11⁩ ⁨months⁩ ago

  Yet another reminder that LLM is not “intelligence” for any common definition of the term. The thing just scraped responses of other LLM and parroted it as it’s own response, even though it was completely irrelevant for itself. All with an answer that sounds like it knows what it’s talking about, copying the simulated “personal implication” of the source.

  In this case, sure, who cares? But the problem is something that is sold by its designers to be an expert of sort is in reality prone to making shit up or using bad sources, while using a very good language simulation that sounds convincing enough.

  source

  • Hyperreality@kbin.social ⁨11⁩ ⁨months⁩ ago

    Meat goes in. Sausage comes out.

    The problem is that LLM are being sold as being able to turn meat into a black forest gateau.

    source

    • brsrklf@jlai.lu ⁨11⁩ ⁨months⁩ ago

      Absolutely true. But I suspect the problem is that the thing is too expensive to make to be sold as a sausage, so if they can’t make it look like tasty confection they can’t sell it at all.

      source
  • CaptainSpaceman@lemmy.world ⁨11⁩ ⁨months⁩ ago

    Soon enough AI will be answering questions with only its own previous answers, meaning any flaws are hereditary to all future answers.

    source

    • samus7070@programming.dev ⁨11⁩ ⁨months⁩ ago

      That’s already happening. What’s more is that training an llm on llm generated content degrades the llm for some reason. It’s becoming a mess.

      source

      • -> View More Comments
  • Fades@lemmy.world ⁨11⁩ ⁨months⁩ ago

    Anyone that needs reminding that LLMs are not intelligent has bigger problems

    source
• MonsiuerPatEBrown@reddthat.com ⁨11⁩ ⁨months⁩ ago

  the ideological composition that:

  if you are allowed to do something then I must be allowed to

  coupled with

  just because i can do something doesn’t mean that you can do it

  … is the basis for all human chauvinism be it gender, racial, or national. And now these fictions … these quasi-legal fictions called corporations … are taking the rights of humans being and laying claim to them while simultaneously declaring human don’t have these rights.

  What the fuck is going on ?

  source

  • MotoAsh@lemmy.world ⁨11⁩ ⁨months⁩ ago

    It’s capitalism, Jim. They can make more profits by stripping humanity from humans.

    source
• tiny_electron@sh.itjust.works ⁨11⁩ ⁨months⁩ ago

  What if they are using OpenAI API and they dont have a model of their own?

  source

  • nightwatch_admin@feddit.nl ⁨11⁩ ⁨months⁩ ago

    Yeah, that was my first thought: Nice, Elon slapped a reverse proxy in front of OpenAI lmao

    source
  • brsrklf@jlai.lu ⁨11⁩ ⁨months⁩ ago

    It’s not what they say happened, and I think people at openAI would not have answered like they did if it was the case. Grok finding answers that users got from open ai and recycling them seems plausible enough.

    Years ago there were something a bit similar when bing was suspected of copying google results. Actually, yeah, sort of : the bing toolbar that some people installed on their browser was sending data to Microsoft, so they could identify better results and integrate them in bing.

    Obviously some off these better results were from people with the bar that were searching on google.

    Someone from google actually proved it was happening by setting up a nonsensical search and result in google, googling for it a bit with the toolbar on, and checking that the same result would then appear in bing.

    source
• ZILtoid1991@kbin.social ⁨11⁩ ⁨months⁩ ago

  Elon be like: "What if I got OpenAI's chatbot and disguised it as mine?"

  source

  • Moof_Kenubi@lemmy.world ⁨11⁩ ⁨months⁩ ago

    Delightfully devilish disruptive, Elon.

    source
• Lophostemon@aussie.zone ⁨11⁩ ⁨months⁩ ago

  Wait until similar code starts being unearthed in Teslas etc.

  Musk isn’t a genius, he’s a thief of IP.

  source

  • Plopp@lemmy.world ⁨11⁩ ⁨months⁩ ago

    “Tesla, open the driver side door.”

    “I’m afraid I can’t do that”

    source

    • XTL@sopuli.xyz ⁨11⁩ ⁨months⁩ ago

      What are you doing, Dave?

      Dave?

      Please stop.

      source

      • -> View More Comments
    • funkless_eck@sh.itjust.works ⁨11⁩ ⁨months⁩ ago

      thats already a feature

      source
    • jetsetdorito@lemm.ee ⁨11⁩ ⁨months⁩ ago

      unless he stole Siri

      “there are no movies in your area by that title”

      source
• xantoxis@lemmy.world ⁨11⁩ ⁨months⁩ ago

  Just scambots eating scams scammed by other scammers, all the way down.

  source
• dojan@lemmy.world ⁨11⁩ ⁨months⁩ ago

  The irony of calling it Grok.

  source

  • Viking_Hippie@lemmy.world ⁨11⁩ ⁨months⁩ ago

    Musk has long since proven himself impervious to irony, responsibility and any expectations of logical and ethical consistency.

    He’s what Nietzsche would probably an “Überidiot” if he was alive today.

    source
• donuts@kbin.social ⁨11⁩ ⁨months⁩ ago

  AI is looking like the biggest bubble in tech history and stuff like this really ain't helping.

  source

  • Draedron@lemmy.dbzer0.com ⁨11⁩ ⁨months⁩ ago

    AI at least has a good chance to become a big thing in some areas. NFTs were the bigger bubble and just a straight up scam

    source
  • bassomitron@lemmy.world ⁨11⁩ ⁨months⁩ ago

    I think you’re underestimating how much AI is used in enterprise. It’s got enormous potential and any tech company ignoring it is just shooting themselves in the foot. ChatGPT isn’t the only type of AI.

    source

    • donuts@kbin.social ⁨11⁩ ⁨months⁩ ago

      I'm don't think I am.

      The internet had a ton of legitimate and potential users too, but that didn't prevent the dot com bubble from bursting.

      Not only is AI built on a shakey house of cards of stolen IP and unlicensed writing, artwork, music and other data, but there are also way too many players in the space and an amount of investment that, in my opinion, goes way beyond the reality of what AI can achieve.

      source
    • RandomStickman@kbin.social ⁨11⁩ ⁨months⁩ ago

      Internet is important and useful, but it didn't stop the .com bubble from being a thing

      source
    • kpw@kbin.social ⁨11⁩ ⁨months⁩ ago

      ChatGPT is the kind of AI that is hyped now. Other kinds of AI (formerly statistics) have been used for decades. Oh you can learn the parameters of some function from the data? Must be AI.

      source
    • banneryear1868@lemmy.world ⁨11⁩ ⁨months⁩ ago

      E-discovery and market simulation tools have basically been using these sort of models for a long time. I think “AI” is a misnomer and more of a branding/marketing term, reserved for the latest iteration of these tools, what used to be called “AI” is given a generic term describing it’s use, and the new thing becomes AI until the next significant improvements are made and it happens again.

      The thing where people think these new language models are going to create a “real” artificial intelligence basically confirms this, it’s almost a religious belief.

      source
• Skeptomatic@lemmy.world ⁨11⁩ ⁨months⁩ ago

  Mistral Dolphin 2.1 said the same to me once. They use GPT-4 for the reinforcement so they don’t have to pay humans, and that sentence must slip in there more than they bother to check.

  source
• rsuri@lemmy.world ⁨11⁩ ⁨months⁩ ago

  I can buy that this was accidental because that answer is way less direct/relevant that what ChatGPT would provide. The guy asked for malicious code and Grok described how to not get malicious code.

  source