The Fruit Salad Therapy Tapes: Tape 116
This Post Was Written By A Human Being
Hello, old friends! Well met! (This is a thing I’ve heard people say, I’m trying it out. So far, I don’t think much of it. Sounds a bit silly to me). Hope you’re all doing well! Apologies for the lack of a Therapy Tape last week - things got a bit busy. I’m back this week and, while I’ve got quite a few thoughts to unpack from the last couple of weeks, in my last post I promised some thoughts on the role of AI within art, and I am nothing (NOTHING!) if not a man of my word, so as promised, here are those thoughts.
The last post was about the ongoing writers’ and actors’ strikes over there in the good old US of, and also, A, with a brief nod towards the specific role of AI within those strikes. This week, I thought I’d expand on that with some more of my own thoughts and, as ever, I’d love to hear what thoughts it sparks for my readers as well!
(Thanks so much to those of you who dived right into the new interactive Chat feature last week - I loved hearing your various hopes and concerns for the future, and hope that the Chat will continue to grow into an interactive community space as both me and my readers grow our understanding of it!)
Anyway, let’s talk about the role of AI in art and storytelling. For a lot of people, I think the consensus on this issue was pretty unanimous for a while - it’s a technology with clear and apparent benefits and applications, most of them limited to the more scientific industries (I’m obviously no expert in science and medicine and the like, but my limited understanding of it would lead me to believe that technology with such in-built efficiency and accuracy would be of huge benefit and poses far fewer ethical issues than it does in the arts, though I may be wrong! I’ve not heard much discussion of this side of things!) But within the arts and creative industries, its use is a clear, insidious error and, more significantly, doesn’t really work. This consensus seemed to be founded on the AI-generated images that people created that had way too many fingers or too many teeth, or the AI-generated “songs/poems/stories in the style of x writer” that all read as peculiarly bloodless and false. Very few of the pieces of AI-generated art that people were bandying around felt like anything other than simulacra, because that’s what they were (whoa, I said simulacra. Really proud of myself, actually). People like Nick Cave wrote thinkpieces on how much AI-generated art sucks, because AI will never be able to approximate the depth of feeling that goes into actually creating a piece of creative expression.
I agreed with this consensus - I was concerned about the emergence of AI in art, because it risks putting swathes of people out of a job, but I was somewhat comforted by the fact that AI-generated pictures or stories or songs are generally easy to spot, and feel like they were written by an algorithm, not a person. In Cave’s piece he’s speaking specifically about the depth of pain and anguish you need to submit yourself to in order to write something truly profound, but I do believe that the exact same philosophy holds true of comedy and other “lighter” forms of art - I think you have to be really in touch with your humanity and your emotional self in order to adequately give voice to the myriad of bewildering feelings that absurdity can give voice to. So I was comforted by the fact that AI art would never really supercede actual human artists, because it could never possibly create anything anywhere near as good.
Then I read this piece in Time Magazine and it changed my opinion somewhat. It turns out that the slightly shonky pieces of art being produced by software like ChatGPT do not represent the pinnacle of current AI technology, and that far more advanced pieces of software are creating work genuinely indistinguishable from something a human might create. I genuinely can’t believe that the short story quoted in that article was not written by a person - it seems to capture so much of the creeping, yawning fear and uncertainty that is inherent in being a person. It’s also remarkable that the story, dealing as it does with a giant, unspeakable and constantly growing existential threat at the centre of a workplace that expands, never commented on, until eventually everyone loses their jobs, is such a perfect analogy for the emergence of AI itself.
The problem faced by the striking writers, then, is no longer that the big studios are obsessed with using technology that increases their profit margins and puts real humans out of work but that ultimately produces inferior results. The problem now is that, in some cases, the AI might actually be producing work just as good as the writers could produce (please note that I am not arguing here that AI is actually superior to a human writer, obviously in my heart I believe that something created by a human will always be superior to something created by an algorithm, but I am trying to get my head around the fact that the argument that “You’ll always be able to tell” no longer seems watertight).
The question here, then, becomes - why do people actually consume art? The studios and big corporations that are actively trying to incorporate AI into their business models clearly believe that ultimately, art is about selling a product to an audience. Quite what that product is doesn’t necessarily matter - whether or not it actually authentically expresses a human feeling, or whether it just resembles the expression of a human feeling, doesn’t really matter as long as it takes more or less the right shape for them to be able to sell it. My own feeling is that I consume art in order to glimpse the infinite different ways in which it’s possible to experience the world. If a piece of art was created by an algorithm, I think it essentially possesses no value. But if I had read that short story while unaware it was written by an AI, it probably would have made me reflect on something, or brought me into contact with an authentic feeling. These studios, then, have access to a tool that makes their business more efficient and more profitable and that audiences can’t necessarily even tell they’re using. It’s a no-brainer for them, if you decide that art is ultimately a business. If you decide it’s something more than a business, then what needs to happen is some sort of unanimous consensus that using AI to make creative work is inherently unethical, because it robs the entire activity of its meaning.
I have no idea if legislation could ever be introduced that actually declares the use of AI to be unethical and forbidden but only in specific instances - declaring it to be fit for use in medicine but unfit for use in any creative environment, for example, could lead to endless wrangling over what constitutes “creativity.” Arbitrarily deciding that there is no place for creativity within science or medicine, and that you can therefore treat those disciplines as entirely distinct, seems very dangerous and restrictive to me.
So the only other workable solution, I think, is to make it a legal requirement that every piece of work made using the input of AI has to self-declare as such, both in the product itself and in all the marketing for the product. Every trailer for a film written by an AI needs to inform the viewer of how it was written. The same for every poster for a book written by an AI. Only then can we find out what the general public actually consumes art for - if all they ultimately care about is consuming something that feels like a story, then people will flock to see these films regardless, and then we have a whole other problem to contend with - how do we continue making art when the general public have made it clear they consider the art they consume to be nothing more than a product for them to buy? But if, as I half-suspect and entirely hope, the majority of people consume art in order to briefly experience the world as someone else has authentically experienced it, then making it a legal requirement for AI-generated art to declare itself as such should, in theory, lead to a series of grand failures.
I learned from your replies last week that certainly not everyone shares my optimism that these strikes will lead to a better future for cinema. I’m not sure my optimism continues over into my feelings about AI. What I’ve described above is a sort of naive hope - that if you were to trial this kind of approach, there would be some fundamental human need for emotional connection that would ultimately lead to the widespread rejection of AI. Sadly, this time I don’t think that is what I really believe. I don’t know what I think the future holds for this, but I don’t think it is likely to be a utopian celebration of the creative impulse. Perhaps the future lies somewhere between that naive ideal and the most pessimistic visions of human obsolescence. Perhaps the real future will just involve people wandering round in a daze, never quite knowing whether the stories they hear and the films they see and the books they read ever came into contact with a real human thought or not. These dazed, confused people will have no idea whether any resultant feelings of joy, sadness, hope, fear etc actually count for anything. “It made me sad, but I forgot to stick around for the credits to find out if it was a person that made me sad or a robot, so now I just feel kind of neutral about the whole thing,” we’ll say, no longer able to differentiate which of our feelings really belong to us and arose in our soul, and which ones arose as the receipt for a transactional exchange with an automated system. It sounds like a future with one more thing to worry about, and I reckon we can all be doing without that.
A New Thing In Comedy - I usually use this section to plug an interesting new show, or film, or whatever, and occasionally have to use it for something sadder. This week, I was devastated to learn that the brilliant comedian Maddy Anholt passed away at the age of 35. There’s a fundraiser organised by her family for her one-year-old daughter, and if there’s anything in the comedy world I’d like to raise awareness of this week, it’s that. Please consider donating if you can.
What’s Made Me Laugh The Most - Katie Norris’s set at the Paddock, about living with a Gen Z housemate. She’s debuting at the Fringe last year, and I think it will be a hell of a show.
Book Of The Week - Just started Yuval Noah Harari’s 21 Lessons For The 21st Century, and can already tell it’ll be one of those books that informs my thinking for a while. Its opening gambit is that humanity has run out of stories to tell about itself and the apocalyptic vibes of 21st century life are down to our inability to find a new narrative for ourselves yet, so we assume our story is simply coming to an end. I’m excited to read on and see if he has any proposed solutions, or if it’ll just be 300 pages of him going “Yeah, so sorry about that, bit of a bummer.”
Album Of The Week - The ArchAndroid by Janelle Monae. This is brilliant! Why didn’t any of you tell me? Other than all those people who kept telling me I should get into Janelle Monae? Thinking about it, it probably is on me that I’m so late to this. I love it when modern pop artists make albums as audacious and over-the-top and genre-busting as the most self-indulgent 70s artists, I think that’s when music feels most exciting. I love this album.
Film Of The Week - Past Lives. This is a really great debut feature by Celine Song about two childhood friends who lose touch when one of them leaves Korea to move to Canada, and their reunion 24 years later, and the myriad different lives they could have led in between if things had been different. I really enjoyed it and you should give it a watch if you can!
That’s all for this week! Do let me know your thoughts, either by replying directly here or adding your thoughts to this week’s interactive chat and, as ever, if you want to send this newsletter to a friend or encourage others to subscribe, I’d hugely appreciate it! Take care of yourselves until next time, and all the best,
PS Here’s a picture of me immediately after winning a game of hide and seek. I went right into the middle of a thicket, and nobody ever checks the thickets thoroughly enough:
Thanks for reading The Fruit Salad Therapy Tapes! Subscribe for free to receive new posts and support my work.