So much art and literature both public and private was used to make these types of LLMs and yet no one truly talks about the consequence of stealing that labor and undervaluing it. So many tech bros love to shit on the arts because a robot can make something that is an amalgam of many artists and call it “art” but forget or willfully ignore the labor needed to create that data beforehand, again it’s labor exploitation just as you mentioned. Also yeah as used as data scrappers and aggregators these types of machines are amazing and something that should be used, it’s just that there needs to be actual human beings making sure the collected data is cleaned and properly organized or else it’s just vomiting out useless shit. The humans that do this type of cleaning work are so well hidden or outsourced to other countries that no one wants to acknowledge that there is no way to truly remove the need for human labor and inputs or else the “magic” of AIs would fizzle out and the current boom would turn to a bust cycle similar to when crypto was so fresh in everyone’s mind.