The sorcerer’s apprentice: The impact of AI on future jobs
- Deryn Graham
What is the future of humans in the world of AI? We look at the myths and realities, risks and opportunities posed by the adoption of AI.
In Disney’s The Sorcerer’s Apprentice, Mickey Mouse employs the services of a broom to help him more efficiently fill a cauldron with water. The broom misses the brief and even when the cauldron is full, continues to bring in bucket after bucket of water. Unable to halt the broom’s advance, Mickey takes an axe to it, but each of the splintered pieces forms a new broom, and soon the place is overrun with over enthusiastic brooms and flooded with water.
This story has been used as an analogy for AI taking over jobs, after humans have used it to improve productivity. Man, it is predicted by AI’s detractors, will ultimately be taken over by machines (or brooms).
Dr Steven James, Senior Lecturer in the School of Computer Science and Applied Mathematics quotes British sci-fi writer Arthur C Clarke who said, “Any sufficiently advanced technology is indistinguishable from magic” and it is this mystification around AI and its uses and applications that is driving the myths around how far it may eventually take us.
“It’s common for most technologies to be overestimated in the short term and underestimated in the long term,” says James. “But AI is showing signs of diminishing returns; having ingested all of the data on the internet, we are starting to see a slowdown in the improvement of these systems. When there is no more data from which to draw, there will be no new outputs, making AI less than intelligent in the human sense,” he says.
Although the displacement of humans in favour of AI, bots and other automated functions may have been hyped, it is certain that, like the internet 40 years ago, new technology has changed the world of work irrevocably. However, it is said that in the next decade, most work roles are not at risk from AI, but from a human who knows how to use AI better than you do. After that, it’s impossible to say where AI may take us. People are still only dabbling in AI and large language models such as ChatGPT, with few fully fledged, scalable systems in commercial use, so keep your eye on your colleague and not necessarily company strategy around AI.
AI is not sentient
Fears around humans being supplanted by AI en masse are mere extrapolations of far-fetched scenarios. People who are surprised by how good ChatGPT and other large language models are, are likely to speculate and imagine ‘what else it can do’. But it is precisely language that sets humans apart from animals – and technology – and AI does not have a human’s nuanced understanding of a question or a prompt. Anthropomorphising technology is a mistake, says James, and ascribing to it human capabilities is simply good marketing on the part of its creators. When we prompt ChatGPT and it takes a moment to respond, it isn’t ‘thinking’, it’s ‘processing’ and studies have shown that using ‘please’ and ‘thank you’ doesn’t necessarily change AI’s output. No one ever used these social niceties with Google.
James believes that in fact, labelling these language models as ‘artificial intelligence’ is a misnomer. They are not intelligent in the way that humans are intelligent and should rather be called a ‘machine learning tool for language’ or a ‘machine learning tool for hiring’, or a machine learning tool for whatever function it is being deployed.
If machine automation has changed the workplace for blue collar workers, mechanising factory production lines, agricultural and other mundane, repetitive tasks, the fear is that it is now coming for white collar jobs. Some believe that this will result in the death of creativity, the erosion of critical thinking and increased job inequality and polarisation.
Bruce Mellado, Professor in the School of Physics believes that AI is simply a tool for improving efficiency and production and cannot be blamed for job losses. “We can’t blame inequality and unemployment on AI or machines, but on an unequal society and an imbalance in economic structures,” he says.
Finding purpose in AI
For Mellado, the critical question is the purpose for which technology is used, which should be for the enrichment of the many and not the few. In mining, technology has replaced some functions, but it can also be used to detect exhaustion in workers, potential hazards and dangers and therefore save lives and so is a good thing. “We need to judge AI on a case-by-case basis, find its responsible application and how and where it can help and do good,” he says. “If the goal is simply profit, there is no doubt jobs will be lost.”
Benjamin Rosman, Professor of Machine Learning and Robotics in the School of Computer Science and Applied Mathematics and Director of the recently established Wits Machine Intelligence and Neural Discovery (MIND) Institute believes that for the many millions of people tied to the yoke of daily work drudgery in return for wages that barely keep them alive, technological advancements offer an opportunity for a societal re-set and review of the meaning of work-life balance.
As AI automates more work functions, Rosman says that this could free manual workers to pursue other opportunities that have the potential to make them more money with a greater sense of personal fulfilment. For example, AI gives people who have not had the advantage of education the opportunity to construct a business model from a modest idea. It can help propel a small business to greater heights. It can model financial projections, enabling people to look more successfully for funding for a business start-up. AI unlocks human potential that has been stymied by social circumstances, bringing more people into the formal economy, creating new jobs. AI gives people the freedom to innovate, and to create a strong gig economy, already part of South Africa’s work landscape.
Work’s existential crisis
Rosman doesn’t believe that there are many jobs which AI won’t be able to do in the future, and this poses an existential crisis for the world of work. AI affects the prospects of all employees, but rather than running off to retrain and reskill ourselves, he says, like Mellado, that it would be better to rethink and reimagine the structure of society and the economy. The expanding uses of AI means that there has never been a better time to realise dreams with fewer resources.
Much has been made of AI taking creative jobs driving Hollywood writers and the Actors’ Guild to go on strike in protest. But AI is not creative insofar as it cannot write with a distinct ‘voice’ shaped by human insight. It is unable to apply different literary devices, wit or subtlety to its output. It’s humans who give nuance as they finetune the work of AI and so it’s unlikely to replace many of the roles that go into producing a movie, a stage play, or even a novel. The Hollywood writers’ and actors’ strike showed the power of labour unions in being part of discussions and contractual negotiations about who controls AI and how it will be used. Labour in all sectors needs to be involved and have input into how AI technologies are deployed, bringing better efficiencies and productivity, cost reductions and improvements to the bottom line without jeopardising jobs. Freeing workers up from performing repetitive tasks allows them to move into bigger, more complex roles, developing a more, not less, skilled workforce.
The question is what we do in an age when there are more resources available but fewer jobs. According to Rosman, we re-think how we derive value from our lives. In the end, it’s entirely up to humans to craft the future we want to see, including the uses and purpose for which we deploy AI.
Sam Altman, CEO of Open AI said, “No one is ever going to be replaced, but jobs are definitely going to go away”. The question is how we use AI to boost innovation and create opportunities while reducing the risks to people’s chances of meaningful employment.
Who controls AI and how it’s used is critical in shaping the future of work, and is something in which multiple stakeholders including government, labour unions, lawmakers, and the public and private sectors need to be involved.
- Deryn Graham is a freelance writer.
- This article first appeared in?Curiosity,?a research magazine produced by?Wits Communications?and the?Research Office.
- Read more in the 18thissue, themed #Work, which delves into the evolving nature of work, shaped by societal shifts, technological advances, and equity challenges.