Books Written Without AI Can Now Receive New 'Human Authored' Certification

Sure, great, but HOW? At the moment at least when a desire is held to profit from written work generated by AI, that desire and motivation comes from a human being. If the Authors Guild wants to confirm that a human being wrote something by basically communicating with that human about the work then they have no way to reliably determine if the human they're talking to generated it by writing down their thoughts or instructing an LLM.

If the quality level from AI work is similar enough to a traditionally written work that the text on its own doesn't clearly indicate machine authorship then the fact of the submission process and the communication between a human being and the Authors Guild could really be the only means by which this is done. So basically, charm them enough and now your AI generated text output could gain extra legitimacy courtesy the Authors Guild because it's now not just you implying you wrote it, it's the respected Authors Guild outright stating it isn't AI.

It also puts in to question some assumptions about this whole endeavour as well. If it's not a quality guarantee, only provenance, as in it can be bad writing but the Authors Guild attests it's bad human writing, then assumptions like "One cannot relate to a bot that does not have its own lived experiences to share" are undermined since that will only hold true on the basis of knowledge the reader has about the text, rather than the text itself resonating with the reader because human generated writing is inherently superior. If that knowledge can be so easily corrupted, it's worthless or at least only a couple of scandals away from being made so. It also gets very messy with things like the example they gave of KC Crowne whose book accidentally included some of the conversation they had evidently had with an LLM while writing the book. It is a hilarious smoking gun that the author used AI tools in the process of their writing, but funny as that is, the mistakenly included text shows that they're at least directing the output and seem to be using the AI to help them refine and make changes to their own writing. They're at least engaging in some form of process beyond simply commanding the machine to generate a book and then selling the result. Defence of 'AI artists' along similar lines to what I just laid out has been sharply criticised and that's pretty justified, right now at least, few would call this idea of directing the output of an LLM, this 'prompt engineering', the same thing as writing, but then again is this a question of degree? Or an absolute? Does the degree to which the author has apparently leaned on this tool affect how much value it has lost to a reader? If the mistakenly included prompt indicates that the author constructed their entire story through prompting, the illusion that the author created this work by synthesising and relating their own experiences is shattered but if it just indicates that they sometimes used it to work through problems while they wrote, is the connection to the author just as sullied, or now only partially? Or not at all? If the Authors Guild accept a submission and put their stamp attesting to its human provenance and later find out that for portions of the text the author consulted with chatGPT to help them work through ideas and test out other approaches are they going to revoke the inclusion in their database? Or is that only if its completely AI generated? In any case whatever answer they have for that can only apply to cases where they know exactly how or if any of the widely available AI tools were used.