[soci]

agreed, extracting tables in pdfs using any of the available openAI models has been a waste of prompting time here too.