Google CEO Sundar Pichai says problems with its AI can't be solved because hallucinations are an inherent problem in these AI tools.

You know how Google’s new feature called AI Overviews is prone to spitting out wildly incorrect answers to search queries? In one instance, AI Overviews told a user to use glue on pizza to make sure the cheese won’t slide off (pssst…please don’t do this.)

Well, according to an interview at The Vergewith Google CEO Sundar Pichai published earlier this week, just before criticism of the outputs really took off, these “hallucinations” are an “inherent feature” of  AI large language models (LLM), which is what drives AI Overviews, and this feature “is still an unsolved problem.”

  • jetAEnglish
    5912·
    6 months ago

    Media needs to stop calling this AI. There is no intelligence here.

    The content generator models know how to put probabilistic tokens together. It has no ability to reason.

    It is a currently unsolvable problem to evaluate text to determine if it’s factual…until we have artificial general intelligence.

    AI will not be able to act like real AI until we solve real AI. That is the currently open problem.

    • JackGreenEarth@lemm.eeEnglish
      153·
      6 months ago

      I think you mean AGI. AI can be as simple as a bunch of if-else chains to win a game of noughts and crosses.

      • jetAEnglish
        135·
        6 months ago

        That was AI has been abused into meaning in the general vernacular I agree.

        By this definition any algorithm whatsoever is artificial intelligence. Including the algorithms Lovelace created before the first computer existed.

        So just like AI used to mean something more than machine learning, AGI will be abused until AGI means the same thing. So I expect journalists to use the appropriate language, or at least explain why they’re abusing language

          • QuadratureSurfer@lemmy.worldEnglish
            31·
            6 months ago

            Sure, but the problem is that our language has evolved and “AI” no longer means what it used to.

            Over a decade ago it was mostly reserved for what you’re describing (which I would call “AGI” now). However, even then we did technically use “AI” for things like NPCs in video games. That kind of AI just boils down to a bunch of If-Then statements.

        • QuadratureSurfer@lemmy.worldEnglish
          3·
          6 months ago

          I think any time “AI” is involved, journalists should be much more specific about what exactly they’re talking about. LLMs, Computer Vision, Generative models (text/image/audio), Upscaling (can start to get a little muddy here between upscaling and generative models depending on how this is implemented), TTS, STT, etc…

          I definitely agree that “AI” has been abused into the definition it is now. Over a decade ago “AI” was mostly reserved for what we have to call “AGI” now.

        • Aceticon@lemmy.worldEnglish
          3·
          6 months ago

          As somebody who uses what has long been called AI in game making (stuff like pathing algorithms and steering behaviours) I would rather we don’t stop calling those things that just because a bunch of greedy assholes are misusing the term for the purposed of getting a bunch of hype-trains going for maximum personal profitabiliyty on the backs of techno-ignorant “investors”.

          I’m still pissed of at how the greedy assholes fucked up the Internet from what it was back in the 90s.

    • brbposting@sh.itjust.worksEnglish
      2·
      6 months ago

      Media is speaking to a nation who voted for a man who bragged about grabbing women by their genitals is almost majority below average. (yes dumb joke)

      Models know how to arrange text far better than millions and millions of people. Is it terribly unfair to condense “artificial, simulated (non-reasoning) pseudo-‘intelligence’” down to “AI”?

      Not for you - is it unfair for the general public?