• ShrimpCurler@lemmy.dbzer0.com
    link
    fedilink
    arrow-up
    4
    ·
    9 months ago

    Might not be an accurate translation but it still speaks the truth. Unfortunately there aren’t many ways of avoiding the future.

  • 1Fuji2Taka3Nasubi@lemmy.zip
    link
    fedilink
    English
    arrow-up
    1
    ·
    edit-2
    9 months ago

    This problem stems from この先 being able to mean “ahead from here in space”, or in time which (mis)translated to “the future”. Without proper context (that it is a sign on a road) the translation software had to make a guess, and it guessed wrong.

    It may be possible to infer from 前へ行く in the second sentence that it is more likely referring to space than time, but I still think it is possible to construct some similar sentences which even humans might misunderstand.

  • hydroptic@sopuli.xyz
    link
    fedilink
    arrow-up
    0
    ·
    9 months ago

    Google Translate is generally hot garbage. I’ve actually found DeepL surprisingly good especially with more “niche” languages like Finnish, although it does definitely sometimes get things hilariously wrong

      • hydroptic@sopuli.xyz
        link
        fedilink
        arrow-up
        0
        ·
        edit-2
        9 months ago

        Is that DeepL’s app? I’ve never used it so I have no idea what the UI looks like and I didn’t try to translate the sign’s text with the web version.

        Edit: out of curiosity I tried out how the web version handles this sign:

        • Bagel5941@aussie.zone
          link
          fedilink
          arrow-up
          0
          ·
          9 months ago

          Yeah, it’s the android version.

          To be fair, It looks like a problem with the OCR from the app rather than the translation. When I use the phone’s native OCR and copy/paste the text into DeepL Translate, I get the same result as you.

          • hydroptic@sopuli.xyz
            link
            fedilink
            arrow-up
            1
            ·
            9 months ago

            Ah right, that definitely makes sense. I can imagine OCR’ing Kanji could be a bit of a nightmare