• tetris11@feddit.uk
      link
      fedilink
      English
      arrow-up
      10
      ·
      edit-2
      2 days ago

      I have a good friend like you. Smart guy, talks about R but mostly about Julia.

      Decided to write his own transpiler. Dropped off the face of the earth. I wish I had more friends like that

      • lonesomeCat@lemmy.ml
        link
        fedilink
        English
        arrow-up
        7
        ·
        2 days ago

        Developer here, I seriously don’t know what to think of the modern tech in general. I reject LLMs because they’re not trustworthy so all of my searches are your typical search engine, stackoverflow, etc. The internet feels so empty, you’d rarely find anything from the last couple of years that’s not AI-gen.

        This is making believe that we will reach a point when our tech stack docs are only cooked into LLMs and I fucking hate that. I would rather go back to the 1990s over this.

        • Riskable@programming.dev
          link
          fedilink
          English
          arrow-up
          2
          arrow-down
          1
          ·
          21 hours ago

          The problem is it’s an arms race! Stop treating LLMs like their existence is the problem and start viewing it via the lens of war:

          The enemy has lots of badly-behaving LLMs! Marketers, scammers, and lazy management are equipped with Big AI brand LLMzookas that are sending hallucinations our way!

          Captain: “So what do we do about it?”

          Soldier: “Captain, there’s FOSS LLMs that we can deploy! We can use them to defeat the enemy’s bullshit slingers! They can be used to search the web on our behalf to filter out hallucinations and advertisements disguised as content! We can set them up to monitor enemy deployments and analyze intelligence to find the truth and stop propaganda in its tracks!”

          Captain: “…but can FOSS AI generate boobs‽”

          Soldier: “Sir, FOSS has already surpassed commercial AI in that front of the war.”

          Captain: “We need to deploy FOSS AI ASAP!”

      • ZoteTheMighty@lemmy.zip
        link
        fedilink
        English
        arrow-up
        4
        ·
        edit-2
        2 days ago

        Jupyter notebook isn’t a language, it’s a tool for running interactive sessions, typically with Python, but in principle with any language. I’m fairly certain people run Julia in Jupyter Notebooks.

        As for the advantages of Julia versus Python, arrays are native types, so they interface better across the entire language. It’s also shockingly fast in comparison, it compiles the code at runtime, so the longer the program runs, the faster it is.

        If there is a use-case for R that Julia or Python can’t do, I haven’t seen it. I personally don’t see the point of writing code in R when the Python and Julia are more broadly useful to learn.