3 ms·

Semi related but, googling a problem just to stumble on an 100% written by AI Medium article as the top result, is the bane of my existence. So really, I can g

• stonethrowaway 2y ago
  Assuming it answers the problem, what’s the issue?
  • tempfile 2y ago
    The answer is probably wrong?
  • sofixa 2y ago
    It wastes your time because it's full of cruft. And you cannot assume to trust it will solve your problem, so you're risking losing even more time.
    • ziddoap 2y ago
      While I'm not about to defend AI blog spam, what you just said it completely applicable to human-generated blogs as well. You should not just assume something is trustworthy or will solve your problem just because it was written by a human. And I have read plenty of cruft-filled blogs/articles/etc. from humans.
      • rsynnott 2y ago
        > While I'm not about to defend AI blog spam, what you just said it completely applicable to human-generated blogs as well. It kind of depends on the subject. If the blog is about politics, say, then you may assume that the human author is likely to be willing to lie. If it's about a non-contentious subject, though, then it is highly unlikely the human author will lie, and if they don't know, then, well, okay, _some_ people will write articles about stuff that they don't understand, but it's not common. Most non-psychopaths, if they don't know the answer, won't just make something up. Whereas the magic robot will very happily spew nonsense on any subject.
      • consteval 2y ago
        If a human has written it I can be somewhat confident the code was at least compiled and tested a bit. The same is not true for AI generated prose, in which the author often can't even be bothered to remove AI boilerplate. In such a case it's obvious the code has never been tested, and it might not even compile.
  • bil7 2y ago
    usually I expect an instructional article/blog post/etc to have actually been tried and tested. If it has the format of "I had x problem, chatGPT suggested solution y, it actually worked, here it is and a bit about why it works" then of course that's fine and good. But you'll only have to google a few times before you see a 100% synthetic SEO optimised useless article that does nothing but waste your time.
  • Vegenoid 2y ago
    This assumption would often be false, which is the problem. I already have access to LLMs (that I leverage often), I’m using a search engine because I’m looking for high-quality, detailed info or real-world examples from someone who knows what they are talking about. This blind cheerleading of LLM-generated content filling the web is what pushes people to hate them.
• ayaen 2y ago
  Some people dont even bother with cleaning up the boiler plate: "Sure, here are 11 examples of recursive islands..." Yeah, that's how some articles begin