• Hamartiogonic@sopuli.xyz
    link
    fedilink
    English
    arrow-up
    2
    ·
    8 months ago

    This is the way.

    Given that there have been signs of the ML industry running out of quality data, there’s a good chance that development will begin to show down. Nowadays, the data is nearly always contaminated with AI generated trash, which means you shouldn’t use it to train a new model. Eventually, we’ll hit a point where it’s nearly impossible to improve the model because you just can’t find the right kind of data for it.