• @Aceticon
    link
    1
    edit-2
    1 year ago

    Well, you can train a Neural Network on pretty much anything, even if there’s no point.

    Putting ML in the middle doesn’t make it any less susceptible to the general problem with mathematical modeling that is Garbage-In = Garbage-Out - I would even say it makes it more susceptible because those doing it don’t even need to begin to understand what’s going on, unlike with more straightforward algorithms and formulas.