Pattern recognition and iterative analysis are valid applications of large data models, but the primary application of regurgitative plagiarism is a novelty at best. At some point we should grapple with the cost/benefit of the extractive capitalist models we are applying to everything now.