• GreenBottles@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      9 hours ago

      I find they fail a lot and spit out code with lots of problems more often than not.

      Studies also show that humans have better quality output and can apply reasoning and logic where LLMs cannot. LLMs will just hallucinate to fill blanks… Its not even close to prime time output.

      Not to mention very insecure code output as well…

      Its not “AI” its hype around text prediction.

      • droopy4096@lemmy.ca
        link
        fedilink
        English
        arrow-up
        1
        ·
        6 hours ago

        This is irrelevant as CEO’s and shareholders keep screaming “moar AI!”. Engineers know limits of the tool but are forced to ignore those. So yes,LLM is comming for your job but not because it can do it,but because it is perceived to be cheaper.