AI is Coming for Your Job – But South Africa Doesn’t Have to Just Watch

While the US is throwing around US$500-billion (that’s R8.1-trillion, bru) at AI, our national AI policy got canned for using *fake* sources. Jislaaik. It's a proper wake-up call, isn’t it? We’re getting shown up by Kenya, Egypt and Rwanda – countries that are actually *doing* something while we’re still busy with kak governance documents. This isn't about robots stealing jobs, it's about who *owns* the robots, and right now, we’re looking at being renters in a world built by others. ## So, How Befok Are We Behind? Let's be real, South Africa is trailing. Badly. The US and China control roughly 90% of the computing power needed for proper, cutting-edge AI, and they apparently own *all* 50 of the top-ranked foundation models. All of them. While everyone else is building, we’re… well, we were building a policy based on AI-generated lies. Embarrassing doesn’t even cover it. It’s like trying to build a lekker braai with wet wood. You’re going to be there all day and end up with a smoky mess. Microsoft’s R5.4-billion investment is a start, sure, but it's a drop in the ocean compared to what’s happening elsewhere. It’s like trying to fill a dam with a garden hose while a monsoon is hitting the other side of the country. ## The R8.1-Trillion Question: Why Money Matters More Than Policies Forget the fancy talk about ethics and regulation for a minute. The real bottleneck isn’t another policy document; it’s the cash. That US$500-billion Stargate project in the US? That’s the kind of scale we’re talking about. India, South Korea and Japan are snapping up graphics chips by the hundred thousand. We’re fumbling with a policy that was built on fabrication. The amount of money being poured into AI globally is staggering – R8.1-trillion is a number that's hard to wrap your head around, bru. It’s more than our entire national budget, probably. This isn't about being clever; it’s about having the muscle to play the game. You can have the best rugby strategy in the world, but if you're getting physically dominated up front, you’re going to lose. Same applies here. ## Data Centres and Eskom: The Unsexy Truth About AI Infrastructure You can’t run an AI economy on candlelight, and let’s be honest, with load shedding being a daily reality, we’re basically running on flickering hope. Data centres – the places where all this AI magic happens – need serious, reliable power. And a lot of it. It’s critical national infrastructure, boet. We already have cloud regions here run by Google, AWS, Oracle and Microsoft. That's a foothold, but we need to turn that into *local* capacity. It can’t just be a transit lounge for foreign platforms. We need to be building, owning, and controlling the infrastructure ourselves. Otherwise, we’re just letting the profits flow offshore, and that’s kak for everyone. ## Taxing the Robots: How to Fund a Future Where Jobs Disappear This is where it gets tricky. If AI keeps shifting income from wages to capital – meaning the owners of the machines get richer while everyone else struggles – our current tax system is going to collapse. We tax wages, and if there are fewer wages to tax, we're in trouble. The answer? Start taxing capital, digital services, and AI-driven profits *now*. Build a solid financial base for any future basic income support. Ideally, we need to do this with other developing economies to prevent capital from simply fleeing to wherever the taxes are lowest. Get this right, and a basic income is sustainable. Get it wrong, and it’ll fall apart at the first sign of trouble. ## Brain Drain is Real: Why We Need to Keep Our Tech Talent Here Egypt graduated 1,300 AI specialists in 2025 *alone*. One. Thousand. Three. Hundred. We're losing coding talent as quickly as we can train it. We need to create real incentives for those graduates to stay here. Better salaries, better opportunities, a more stable future. An AI economy with no engineers is a customer, not a competitor. It’s like having a Ferrari with no driver. Pointless. ## Playing Our Hand: Leveraging BRICS and African Partnerships We’re not going to win this fight alone. We need to leverage our position within BRICS and the African Union. Chinese AI models are already cheaper and becoming more widespread across the Global South. This gives us some leverage, provided we negotiate as a bloc for affordable access to models and hardware on terms *we* help set. Even a small sovereign or open-source capability is better than being completely reliant on a single supplier. ## A Safety Net for the Real World: Basic Income and Retraining AI is going to disrupt the job market. That’s a given. We need a safety net for the people who get left behind. A targeted basic income, pitched at or near the food poverty line, isn’t a utopian dream. It’s an achievable floor. Coupled with retraining programs and support for small businesses, it’s a realistic way to protect people from ruin. The hope is that if AI drives down the cost of production, the price of essentials like food, transport and electricity will fall too. A basic income could then actually buy a decent life, not just bare survival. But that depends on those price drops reaching the people who need them most, and on the productivity gains being shared, not hoarded by the owners of the machines. Look, it’s not a technological gift. It's statecraft. Deliberate, early, and unglamorous. Get it right, and South Africa can catch a fraction of the upside and build a safety net for those who are displaced. Get it wrong, and we’ll be importing our intelligence, our robots, and our inequality, and paying for all three. So, is a basic income the answer? Or are we just delaying the inevitable? Click here to find out what economists are saying about the future of work in South Africa.