Raw LLM Responses

I wouldn't even say it is more like glorified auto complete. We just don't know. Because all of these companies making these claims **refuse to provide proof.** Open Source projects are typically extremely hostile to any AI usage, so we can't draw much from the important ones. But there is absolutely a reason why Meta, Microsoft, Google, and all the others who have open sourced some projects aren't highlighting "these commits are from AI... these modules were never touched by a human" etc.. They are almost surely fucking around with wordplay. Think of what companies like Ubisoft do as they try to turn a string of failed products into salvageable wins by reframing success from "x millions of players bought this game" to "we had x million *impressions*" without further defining what they consider to be an impression. If all these companies, with extreme financial interests in AI surging, were truthful then they wouldn't be hiding the proof of these claims.