• OpenStars@discuss.online
    link
    fedilink
    English
    arrow-up
    4
    ·
    5 months ago

    That’s precisely what I was thinking, but reflecting more on it, I don’t know how well it would handle the webpages, so maybe some other languages mixed in too (I’m out of date, maybe PHP?). If AI writing code worked it would lower the barrier, but I’m not certain we’re quite there yet to trust anything it would create.

    • GBU_28@lemm.ee
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      5 months ago

      Python web scraping is just fine, with the llms you.have the option of either extracting the html and having the LLM read.over that, or having a vision ai OCR the page and make its own decision of what to extract.