A feature of the GAI discourse which I find particularly frustrating is the tendency to counterpoise the efficient and instrumental writing produced by LLMs to the imagined freedom of human authoriality. There are some commentators who seem genuinely shocked by the prospect that the world will soon be full of mediocre writing produced to serve an instrumental purpose. As much as I love Nick Cave he captures this attitude perfectly in this early response to ChatGPT:
ChatGPT is fast-tracking the commodification of the human spirit by mechanising the imagination. It renders our participation in the act of creation as valueless and unnecessary. That ‘songwriter ‘you were talking to, Leon, who is using ChatGPT to write ‘his’ lyrics because it is ‘faster and easier ,’is participating in this erosion of the world’s soul and the spirit of humanity itself and, to put it politely, should fucking desist if he wants to continue calling himself a songwriter.
ChatGPT’s intent is to eliminate the process of creation and its attendant challenges, viewing it as nothing more than a time-wasting inconvenience that stands in the way of the commodity itself. Why strive?, it contends. Why bother with the artistic process and its accompanying trials? Why shouldn’t we make it ‘faster and easier?’
https://www.theredhandfiles.com/chatgpt-making-things-faster-and-easier/
There is a horror here of the idea that cultural production could be mechanically instrumentalised which reflects (in part) the everyday experience of a man who has enjoyed absolute autonomy as a cultural producer for decades. The point in raising this is not to disagree with Cave per se, I do think GAI by necessity will tend to remain at the level of pastiche or collage, leaving us artistically impoverished if we normalise it as a means of creation. This is one reason why I emphasise the importance of using conversational agents as interlocutors rather than producers. The aesthetic, intellectual and epistemic value of a conversation you co-produce with them is distinct from the value which can be found in something you rely on them to produce under instruction from you.
Why does this songwriter want to work in a way which is ‘faster and easier’? What benefit could that bring them? Is it simply because he fails to see he is “participating in this erosion of the world’s soul and the spirit of humanity itself”? Is the solution to this problem to restrict the dignity of the category ‘songwriter’ (see also ‘writer’, ‘thinker’, ‘intellectual’, ‘author’ etc) to those who take a principled stance of refusing to do this*? Much as I think Cave’s horror, understandable though it is, needs to be seen in the context of his own success and the autonomy he enjoys as a result, the willingness to lean on LLMs in problematic ways needs to be similarly contextualised. The reality is that “writing at others’ bidding” as the linguist Naomi Baron puts it, is the norm rather than the exception.
What makes academics want to publish more? I write because, as Baron puts it, “there are things I want to think about and share with others”. But I also write because I periodically have to submit CVs which show documentary evidence of what I spend my non-teaching time at work doing. I write because it’s the most immediate way to influence my field and discipline, establishing myself in terms of the hierarchies which structure the academy. I write because I’m paid in part to so and if I stop my employer will eventually notice and tell me to start again**.
This can feel like a split existence in which the imagined freedom of the scholar uneasily coexists with the mundane reality of being a knowledge worker within an educational bureaucracy. There are spaces of freedom and joy which can be found at this intersection, even if they are fleeting and precarious. But starting from the assumption of an intellectual autonomy which is, ultimately, a fantasy is rarely conducive to finding and dwelling in these spaces. I’m not suggesting we abandon our creative ideals but rather that we see them as challenges to be sustained in a practical register, rather than ideals against which the reality of our employment (or lack thereof) will inevitably show up as diminished and unsettling.
This matters for LLMs because the split I’ve described makes it far too easy to say one thing then do another. To congratulate yourself on your critical stance then go back to reproducing academic practice in an entirely conventional way. If there is an ontological gap between your creative ideals and the reality in which you (fail to) pursue them, it’s easy to adopt short term fixes in relation to instrumental demands while reassuring yourself that this will help get you to a point where you can avoid doing this in future. In imagining ourselves as precious fonts of creativity we disavow the many ways in which our creativity is compelled and impelled.
In a sense I agree with Cave that there is an “erosion of the world’s soul” every time an academic uses ChatGPT to spit out some text which they then lazily regurgitate in a piece of writing. I just think this moral critique needs to be directed at the system which incentives hyper-instrumentalised writing in a competitive spiral, rather than the individual writer themselves. The moral condemnation doesn’t help us navigate the practical dilemmas we face in finding ways to square the circle between writing as a passion and writing at others’ bidding. If we can only write out of passion when we are entirely free from instrumental considerations, a beautiful soul freed from the compromises of the world, we will rarely find pleasure in writing under the circumstances in which we are doing it. In which case why not let ChatGPT do it for us?
*This is one I would increasingly support. Naomi Baron raises the prospect of “Sign[ing] a pledge that written work bearing your name is your own”. I think this fails to grapple with the complexity of what “your own” means but I broadly agree with the sentiment.
**These pressures are even more pronounced if you’re precariously, under or unemployed as an academic while pursuing secure academic employment in the future.
Mark Carrigan 