AI models need retraining regularly to avoid drift, so how are they gonna do that now that the internet is saturated with slop

this whole thing looks like a hail mary. they thought the dataset was big enough to produce AGI, but they were wrong, and now the dataset is permanently poisoned