Five years ago, President Cyril Ramaphosa asked South Africans to imagine something amazing: a brand-new, hi-tech “smart city” for the modern age. “Has the time not arrived to build a new smart city founded on the technologies of the Fourth Industrial Revolution?” he asked.
At the time it felt like a big, exciting idea. But today that idea feels like a distant dream we all had that never came true. In that regard, when it comes to technology advancement, SA is moving slowly. As for AI, the country seems to be going nowhere fast. The government, which should be leading the charge, has been left far behind.
Think about what AI could do for the country. It could help doctors spot diseases earlier. It could help police prevent crime before it happens. It could ensure social grants reach the rightful beneficiaries and stop fraud in its tracks.
The potential for efficiency that AI presents is huge. But so far implementation has moved at a snail’s pace. Why is this so? The answer is simple: the government lacks the funding, necessary skills and basic digital systems to make it work.
Just look at this week’s news. The agency that pays social grants, the SA Social Security Agency, has announced that it will start using biometrics — fingerprints and facial recognition technology — to combat fraud. While this is a good step, it’s a basic one. It’s like finally getting a dial-up internet connection when everyone else is using fibre.
What the government is providing isn’t the advanced AI we were promised; it’s just playing catch-up with tech that’s already years old. The real problem is that our government computer systems are outdated, slow and can’t even talk to each other. Trying to add modern AI to this mess is like trying to put a self-driving car engine into a 1980s’ minivan. It’s a huge, expensive job that has to start from scratch.
Five years ago a special presidential commission handed the government a detailed plan in the form of an “urgent report” that said “the future is now, and we need to act”. It contained brilliant ideas, including investments in training people with the right digital skills. The plan also included creating a special AI Institute to drive innovation in health, farming and energy.
It has been five years since the “urgent report” was tabled. What do we have to show for it? Just about nothing. The AI Institute is barely a reality. We are not training an army of tech-savvy graduates. The University of Johannesburg stands as a lone institute offering a dedicated undergraduate AI degree. Other institutions offer postgraduate and short courses in AI.
Trained AI technicians are being hired by private companies. Others move overseas for better jobs. The government can’t compete in hiring these AI technicians because of low pay and slow and complicated hiring processes. But the biggest reasons for the lacklustre advancement of AI are even more basic: AI can’t run without fibre internet and reliable electricity supplies.
The government is nigh on bankrupt and is struggling to deliver services to its citizens. In that situation, spending money on futuristic AI projects feels like a luxury. It’s a “nice-to-have” when there are so many “must-haves” that need funding first. There are, of course, islands of excellence, primarily in the private finance sector, where necessity drives innovation. But a nation cannot thrive on these islands alone.
Ramaphosa’s smart city was meant to be a hub of innovation. Instead, it remains a ghost town in a speech, a symbol of a promise broken by a state that cannot overcome its own deep-seated contradictions. Until it does, SA will remain a spectator in the AI revolution — if it can even log on to the video feed with its outdated hardware and slow internet connection.
• Lourie is founder and editor of TechFinancials.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.