Anybody who had their ear to the ground regarding Artificial Intelligence for the last year – and many who didn’t – will know that the technology has improved at a dramatic and frankly startling pace, and in many domains we least expected. AI generated art – by tools like Dall-E and Midjourney – evolved from a primitive toy more akin to a child’s drawing to something fully capable of simulating a vast variety of art styles in remarkably persuasive ways. The latest step has been taken by OpenAI’s ChatGPT, a deep learning AI that on the surface purports to be a ‘Chatbot’, but has been widely used since its release as a remarkably competent assistant for a multitude of tasks – including writing code, to the simultaneous delight and alarm of programmers everywhere. Indeed, writers of all sorts, including bloggers, have reason to pay attention, as the artificial intelligence could generate an article in seconds – and to be honest, with greater aptitude than I’m capable of most days.
While many professionals are right to worry that their employers will automate their jobs, they should also be aware that these tools offer them great opportunities for their own work as well. With the help of these AI tools, artists, programmers and writers, among many others, all have the chance to vastly enhance their own abilities – to do in a day what might have taken them weeks. Digital artists know well that tools such as Photoshop make countless tasks easier, mathematicians know that calculators save hours of onerous and demanding work. New and improved tools allow users to set their sights on far more ambitious projects. To learn, to iterate, to experiment. Perhaps you wanted to make a game, but lacked the resources to hire an artist, a musician, a writer, and a programmer? We think about the power it gives large businesses, but sometimes we neglect the power it gives anybody the ability to make something incredible.
Keeping one eye on the discourse, I’ve seen a mix of jubilation and deep concern. Having been involved for a few years in the art scene as I performed the often gruelling task of learning to draw, I was exposed to a lot of the concerns artists have for AIs such as Midjourney. Deep learning algorithms are trained on vast amounts of data, much of which the developers did not ask permission to use, and the outputs often show more than a passing similarity to the data it was trained on. There was, I think, a rightful sense of indignation among artists, especially professionals, who not only had their jobs to worry about, but felt that their own hard work had been stolen to create the machine that would render them obsolete. There is also a valid concern that as machines continue to take over, the opportunity for craftsmanship, and the subtle joys it provides, diminish. At the very least, the nature of craftsmanship changes.
One does not need to look back far, in a historical sense, to have seen these trends before. ‘Luddite’ is often used as a derogatory term now, but many Luddites during the industrial revolution were rightfully concerned that the industrialization and mechanization of what used to be skilled labour would render their years of training obsolete. Craftsmen were largely (but not entirely!) replaced by factories that could mass produce shoes, or tools, or furniture. Now AI will be doing the same for a new vast contingent of professions that people have spent their whole lives becoming proficient in – that people have gone into vast debt to learn, on the promise of gainful employment. As our machinery grows ever more sophisticated, so too does it ability to perform jobs that were once the domain of skilled humans, humans who had spent their lives developing these skills. More cheaply, more efficiently, more easily.
It also brooks reminding that the Industrial revolution, for all of the plenty and prosperity that it produced, also forced a chaotic restructuring of society which, in no small part, created squalor, misery, revolution and civil war. The rise of communism was a direct product of concerns over rapid industrialization and its human cost, and in Russia and China alone millions died as these revolutionary states industrialized themselves. War at an industrial scale produced the horrors of the First and Second World Wars, and the misery increased every bit as fast as the throughput of bullets and bombs. We may look on the fruits of industry today and sense that these advances in technology are good, but we must also be aware that there is a destructive element to any great change to our society.
Technology is a force of nature. Just like the societies who rejected agriculture, or firearms, or industrialization, or any other revolutionary technologies were overcome by those who embraced them, trying to fight AI may be a losing battle. Perhaps AI is just the latest phase in the “Computer Revolution” which has been going on for decades. Perhaps the computer revolution is just the latest phase of the industrial revolution that preceded it. The immensity of what we have accomplished, the sheer power to alter our lives for better or worse, marks one of the most significant technological revolutions in human history. The question is – is the ultimate outcome of the computer revolution that humans are rendered entirely obsolete?
Perhaps, perhaps not. Anybody who tells you with certainty they know the answer is lying to you, or lying to themselves. There can be no certainty – the world is a vast, complex system that is deeply unpredictable. But anybody who tells you there is nothing to worry about is also naive. We would do well to tread lightly. Rejecting AI will not work, but we must be careful in how we implement it. For every factory that made cars, computers and stoves, we had factories that made tanks, and bombs, and poison gas. AI will offer ample opportunity for creation and destruction as well. We might not be able to swim against the current, but we had best hope that we are capable, as a society, as a species, of directing it wisely.