Comment on Unmasking AI’s Role in the Age of Disinformation: Friend or Foe?

• hendrik@palaver.p3x.de ⁨5⁩ ⁨hours⁩ ago

  I'm not a machine learning expert. But I think it's just that we haven't yet learned how to do it yet. It's not a technical matter, or the question where to put it. But more that science has to figure out a few things. It'd be massively useful to guide these things. To control whether they hallucinate or tell the truth. To make them just do customer support based on factual information instead of also engaging in intimate cinversations. To strip bias and stereotypes. To make them "safe". But if you look at these systems in practice, you'll see they often fail. And then someone writes a news article every few weeks. And it happens to all current AI systems, even the market leaders. So I figure science just can't do it yet and we're in the early stages. Nobody knows at this point where to put it so a company could trust AI to act exactly in their interest. We might be able to do so, but that's still science fiction undtil we arrive at Skynet.

  source

  • Lenient_vegetable@lemm.ee ⁨5⁩ ⁨hours⁩ ago

    Do you think they have unreleased more sentient and less llm in the labs or do you think we arent near that level of tech yet.

    I often wonder what it would be like if they stuck it into the quantum chip google has or have they already tried it.

    source

    • hendrik@palaver.p3x.de ⁨4⁩ ⁨hours⁩ ago

      I -personally- don't think so. I also read these regular news articles, claiming OpenAI has clandestinely achieved AGI or their models have developed sentience... And they're just keeping that from us. And it certainly helps increase the value of their company. But I think that's a conspiracy theory. Every time I try ChatGPT or Claude or whatever, I see how it's not that intelligent. It certainly knows a lot of facts. And it's very impressive. But it also fails often at helping me with more complicated emails, coding tasks or just summarizing text correctly. I don't see how it is at the brink of AG, if that's the public variant. And sure, they're probably not telling all the truth. And they have lots of bright scientists working for them
      And they like some stuff to stay behind closed curtains. Most likely how they violate copyright... But I don't think they're that far off. They could certainly make a lot of money by increasing the usefulness of their product. And it seems to me like it's stagnating. The reasoning ability is a huge progress. But it still doesn't solve a lot of issues. And I'm pretty sure we'd have ChatGPT 5 by now if it was super easy to scale and make it more intelligent.

      Plus it's been 2 weeks that a smaller (Chinese) startup proved other entities can compete with the market leader. And do it way more efficiently.

      So I think there is lots of circumstantial evidence, leading me to believe they aren't far off from what other people do. And we have academic research and workgroups working at it and publishing their results publicly. So I think we have a rough estimate of what issues they're facing and what AI progress is struggling with. And a lot of those issues are really hard to solve. I think it's going to take some time until we arrive at AGI. And I think it requires a fundamentally different approach than the current model design.

      source