The writer encountered sleek self-checkout pods at her local Uniq store. Photo: supplied 
Earlier this week, I picked up a jersey from the Uniq store at my local mall. It was quick – grab, swipe, go. But as I went to pay, I was reminded of the first time I walked into a Uniq store and noticed something was different: there were no cashiers. No tills. No queue snaking towards a human behind a counter. Instead, four sleek self-checkout pods stood in one corner, humming with quiet efficiency.
The checkout process at Uniq is simple: grab a paper bag, drop it on the scanner zone and toss in your items. No need to hold up each tag – a hidden barcode reader instantly picks up every barcode and rings up your total in a flash. The screen asks for your rewards card and then prompts you towards the card machine.
Slick. Seamless. A little bit like magic.
The first time, I walked away thinking the future is here. What I didn’t think about – at least not immediately – were the people. The cashiers that might have once worked those tills. The retail staff whose roles had been trimmed down to make way for this whisper-quiet efficiency.
It’s easy to marvel at the tech. What’s harder to spot is what’s been quietly removed to make way for it.
South African workplaces are changing. Not just in the visible ways (more remote meetings, fewer receptionists, less paper clutter), but in the kinds of work we’re doing and how we’re doing it. Spend enough time in offices, hospitals, call centres or warehouses and you’ll see it for yourself. Software is doing what people used to do: algorithms are reading CVs, chatbots are answering customers and machines are filing invoices.
And it’s not just happening in tech startups or Silicon Valley-esque coworking spaces. It’s happening in banks, factories, clinics, farms and even your local Uniq store. The robots aren’t coming – they’re already here. They just don’t look like what we thought they would.
Disruption creeping, not crashing
The conversation about artificial intelligence (AI) and jobs has long been polarising. For every utopian claim that AI will unlock new kinds of meaningful work, there’s a dystopian fear that it will wipe out jobs and leave millions behind. Both are technically true, but neither tells the full story.
A recent McKinsey & Company report, The Future of Work in South Africa, paints a picture that’s neither apocalypse nor fantasy. It estimates that AI, machine learning and advanced robotics could displace 3.3 million jobs by 2030.
But – and this is the bit that rarely makes the headlines – this same report projects the creation of 4.5 million new jobs in the same timeframe. That’s a net gain of 1.2 million jobs. However, there’s a catch: we don’t fully know what these new jobs will be.
It’s worth emphasising that the McKinsey report is a great document and a lot of research went into it, but the numbers it puts forward are still based on speculation. We have no real way of knowing the exact number of jobs that will be gained or lost. McKinsey is making a (highly) educated guess, so keep that in mind as you read on.
Many of these new roles are expected to emerge in the tech space, but they won’t all be obvious tech jobs. They may be hybrid roles like AI ethicists, automation coaches and data-enabled health workers. These are jobs that don’t exist yet, but they will soon seem essential.
According to the report, some sectors will see particularly strong growth. Healthcare and social assistance could (again, educated speculation) add more than 570,000 jobs by 2030, driven by South Africa’s growing and ageing population. Construction, buoyed by urbanisation and infrastructure needs, is projected to create more than 260,000 jobs.
But the gains won’t be evenly spread. For every sector on the rise, there will be one in decline. Retail is expected to shed about 334,000 jobs as cashiers, floor staff and store managers are increasingly replaced or supplemented by smart kiosks and inventory-tracking systems.
Administration-heavy sectors like government and support services may lose 309,000 jobs. Manufacturing, once seen as the engine of industrial employment, is on track to lose 231,000.
Transport, warehousing, agriculture and even real estate are also on the chopping block, albeit to a lesser extent. The information sector, ironically, is projected to lose relatively few jobs (about 7,000), likely because it’s the one designing the systems replacing everyone else.
Not full replacement, just fewer hands
It’s tempting to think of job losses in binary terms: humans out, robots in. But the reality is messier. Very few jobs are 100% automatable. In fact, according to McKinsey’s analysis, 60% of jobs globally have at least 30% of tasks that could be automated.
In South Africa, data-heavy roles such as payroll officers and transaction processors are among the most vulnerable because about 72% of their work activities could be handed over to machines. But this doesn’t mean their roles disappear entirely. It just means fewer people will be needed to do the same amount of work.
In times like these it’s useful to remember that this isn’t the first time technology has rewritten the job script. When the personal computer arrived in the early 1970s it killed off a host of routine clerical roles, from typists and data entry clerks to messengers.
But it also sparked a wave of new opportunities, creating hitherto unimagined jobs for programmers, designers, information technology support staff and network engineers.
The AI shift feels similar, but there is a big difference: it’s faster, more complex and harder to predict. The PC replaced typewriters whereas AI can replace thinking – at least some parts of it.
The challenge that AI brings to Africa is access. AI development is expensive. It demands computing power, data infrastructure and deep technical expertise – in other words, things that are not evenly distributed across the globe.
In regions like Africa, where basic connectivity and skills development still lag behind, the AI era could deepen inequalities rather than dissolve them.
If AI becomes another exclusive tool in the hands of a few, it risks turning into a control mechanism rather than an empowerment engine. In a country already grappling with high unemployment, economic exclusion and vast digital divides, that’s no small risk. It’s the defining question. Governments will need to prioritise getting their citizens high up on the digital curve if they expect them to be able to participate in the economy of the future.
Preparing for jobs with no names yet
We can’t fight the future with nostalgia. There are a lot of things we don’t know about AI in the workplace yet, but one thing we know for sure is that legacy job models are being eroded. What comes next depends on how we respond through policy, education, upskilling and public-private collaboration.
South Africa’s response to AI needs less panic and more preparation – not retreat, but reinvention.
As I stood at Uniq, waiting to pay for my jersey, the lady in front of me approached a self-checkout station with trepidation. She looked to her left then her right, hoping to get the attention of a human being. In a moment, a staff member unpacking stock nearby spotted her and came over.
I could see the tension leaving the lady’s shoulders. The staff member helped her with a smile, putting a bag on the scanner and folding her items into it gently. “Thank you so much, young man,” I overheard her saying to him, “these machines make me so nervous”. DM
This story first appeared in our weekly Daily Maverick 168 newspaper, which is available countrywide for R35.
