• tigeruppercut@lemmy.zip
    link
    fedilink
    English
    arrow-up
    18
    arrow-down
    1
    ·
    3 days ago

    The real question is why anyone would want to use more power than a regular search engine to get answers that might confidently lie to you.

    • Boor@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      edit-2
      18 hours ago

      Google processes over 5 trillion search queries per year. Attaching an AI inference call to most if not all of those will increase electricity consumption by at least an order of magnitude.

      Edit: using their own 0.24Wh number, that equates to 1.2 billion kWh per year, or about the equivalent of 114.3 million USA homes.

    • mrgoosmoos@lemmy.ca
      link
      fedilink
      English
      arrow-up
      8
      ·
      3 days ago

      if it’s Google that they would use us the search engine, search results are turning to shit. it just often doesn’t show you the relevant stuff. The AI overview is wrong. Ads sometimes take up the entire first page of results. so I see why someone would just want to show a question into the void and get a quick response instead of having to sort through five crappy results, after filtering that down from 15 possibly relevant ones

    • TankovayaDiviziya@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      1
      ·
      edit-2
      3 days ago

      I use DuckDuckGo. I use its AI features mainly for stock projections and to search for information on company earnings release. Because when I try to search for earnings schedule by myself, I get conflicting information. DDG AI is actually pretty useful to read troves of webpages and find the relevant information for me in that regard.