I spent the best part of the 2010s working in new media, which – if you enjoyed being repeatedly laid off and then being inundated with jeering messages inveigling you to “learn to code” because your industry was doomed – was a great big laugh. Eventually, the fun began to wear off and in an act of subversive defiance (or cowardly resignation), I took their goading advice, learned to code and pivoted to what I’d hoped would be a far more secure career in “web development”, only for recent advances in AI to supposedly render coding jobs a waste of time, too. It seems I have accidentally timed my career change to coincide with a mass rollout of AI chatbots that have also learned to code, and that are – in many respects – already far better at it than me.
Code can appear alarming to the uninitiated: inscrutable “languages” that mostly read like a calculator having a stroke, but, according to AI’s most fervent evangelists, they no longer need represent any barrier at all. Why bother wrapping your head around the needlessly convoluted nerdspeak required to display white text on a black background, when you can now simply ask a chatbot to do this in layperson’s terms and it will promptly serve up your code, complete with instructions?
Playing around with various chatbots, you’ll still experience the AI making a fair amount of mistakes – which a working knowledge of code helps correct – but you can also just talk these through with the AI, and it will attempt to solve them for you. It’s not difficult to envisage a not too distant future where they can discern users’ needs and walk them through solutions, the role of a human developer seemingly consigned to history.
It’s tempting to succumb to the fatalism around AI job theft here. The technology’s loudest cheerleaders are themselves the most eager to cultivate it, encouraging us to surrender to a robotic new dawn, where devoting the time to learn skills, perform tasks or know about anything may as well be considered a thing of the past. But it fundamentally confuses the ability to shortcut how to do something with developing a full understanding of why you would.
AI chatbots haven’t broken some omertà around coding. They have simply digested a load of resources and open-source materials that were already made freely available online for human beings to learn from. A user could attempt to skip this phase by leveraging a chatbot’s grasp of this knowledge, but in doing so, would forfeit ever grasping what decisions the machine was making on their behalf, why it was making them, whether they were even any good and, crucially, what else was possible.
One of the most gratifying aspects of web design and development is the lateral thinking involved. There is rarely ever an objective and singular correct way to go about achieving something. You have to consider all the different contexts a user will encounter on your site, how you want them to interact with it, what you want to elicit in them, whether what you’ve put together will make their phone immediately overheat and explode, killing them instantly, and so on. A machine trained to aggregate and condense the entire web to its most predictable forms doesn’t think like this, and neither does a user reliant on one.
I’ve been fortunate enough to carve out a practice working on projects where I have been hired not just because I possess coding knowledge the client doesn’t, nor for my creative ideas, but for the combination of the two, and how they inform one another. As well as having been professionally fruitful to learn to code, I also – regrettably, given how I was browbeaten into it – actually … enjoy it? The rush when a harebrained idea improbably works is like nothing else. I unironically think and say things like: “I believe in the capacity for the browser to be an absurdly creative and innovative medium”. There are projects I endeavour to make, whether someone pays me or not.
And while AI may diminish certain aspects of my earning power, but I’m not about to have such a low opinion of my craft that I think it’s essentially the same as typing a command into a chat box. Nobody should.
And yet there’s a concerted effort on the part of Silicon Valley to make us believe that the human mind is predictable, replicable and unsophisticated, and that the arts and adjacent sectors are reducible to a set of equations and keywords, because they’ve spent billions creating machines that can now knock out forgeries of creative endeavour and mildly amusing images of Harry Potter characters wearing Balenciaga.
Asked about potential use cases for AI, OpenAI (creator of ChatGPT) co-founder Greg Brockman gave a revealing prediction about what he saw as the future of entertainment. “People are still upset about the last season of Game of Thrones, but imagine if you could ask your AI to make a new ending that goes a different way, and maybe even put yourself in there as a main character.”
The ability for people to do this has already existed since time immemorial, inside of their own heads. It speaks to such a paucity of imagination on the part of AI’s exponents that they’re asking us to imagine having an imagination. Such people cannot conceive of deriving enjoyment and gratification from creating art, or why someone would prefer to craft their own stories instead of outsourcing the entire process to a machine. They lack even the basic conviction in their own ideas to come up with Game of Thrones fan fiction without asking a computer to do their homework.
The heaviest salivating around the potential of AI is coming from those who see it as an exciting costcutting measure that might allow capital to finally become unshackled from its old adversary, labour. It’s absurd nonsense to suggest that humanity’s collective recorded cultural output has been essentially concluded, bottled and corked at this specific point in history, that it was all merely fodder and data points for training AI models which will take it from here, cheers.
Ingesting every piece of art ever into a machine which lovelessly boils them down to some approximated median result isn’t artistic expression. It may be a neat parlour trick, a fun novelty, but an AI is only able to produce semi-convincing knock-offs of our creations precisely because real, actual people once had the thought, skill and will to create them.
The spectre of AI will be used as a threat and a cudgel by those who see creative pursuit as only possessing worth if it can be monetised, but they’re wrong. A machine has no capacity for self-expression, no compulsion to communicate: this is who I was, this is how I felt and this is what I stood for. We do, and in all of our endeavours we need to start refusing attempts to make us forget how valuable our humanity really is.
Tristan Cross is a Welsh writer based in London
