AI is automating laziness and ignorance – which is wild in SA
· Citizen
The streets of South African cities are wild, but there’s a reason you don’t see 13-year-old children rolling down an inevitably named Mandela Road in a forklift, carrying plutonium-239, selling their own made moonshine to drivers passing by.
While there still remains some semblance of law in the country, there are certain tools that require a degree of knowledge and sophistication to operate. If you disagree, let’s go find an operating table. I know somebody with half a music degree who could fix your brain.
Visit betsport.cv for more information.
AI requires skills
AI shouldn’t be an exception. With a forklift, at least the damage can be geographically contained, but with AI? You could pull the wool over the eyes of the entire state… and we saw it happen.
The Cabinet approved the draft national AI policy on 25 March. There must have been a very cute cat at the printers because it only got gazetted on 10 April. It then took over two weeks to get withdrawn. We know why it was withdrawn; being replete with AI hogwash itself, but what happened in the preceding two weeks is what’s most scary.
The race for relevance saw legal experts, academics and AI consultants flooding social media with announcements of this policy and their thoughts on it. You’d think that an expert reading a draft national policy would notice some citations seemed off, but it doesn’t seem so. Experts were interviewed on the news talking about this policy, and none of them said: “Hey, there’s something in here that seems fishy”. No, instead, South Africans had to rely on journalists to point out the stupidity.
It took so long for even the experts to realise their lacking expertise that one must really ponder how much expertise we collectively have and whether we’re responsible enough to wield it.
Who’s responsible?
But that’s going to require too much thinking because when it comes to AI, the government’s current priority seems to be to keep on using it to stuff up. We now have Minister Khumbudzo Ntshavheni referencing a non-existent report to tell us how misinformed we are. This in an attempt to tell us that social media must label AI-generated content.
We used to have architects and engineers draw building plans on paper and awesome-looking angled desks. With the dawn of software that could rid them of the desks and allow them to draw on computers, we didn’t suddenly also remove the requirement of learning to draw. We didn’t say that the computer can do the load calculations for you, so don’t bother with math. We didn’t say, outsource your skills to the press of a button. We understood that there would be outcomes and somebody would have to be responsible for them.
Nothing has changed with the inception of AI. Only we’ve given these tools to people who can scarcely read and expect them to be able to account for the outcome?
How would automating somebody’s laziness and ignorance be any good for the country? It’s too convenient to start getting rid of it now that everybody has had a taste of outsourcing their lack of ability.
If the new skill in demand is going to be writing prompts, then we’re already in trouble because writing is already a luxury in the skills arsenal of 2026. No amount of policy direction and dictation is going to change that.
Instead of trying to dance around how we should police AI, we should just do what we should have done in the first place: hold the people who put stuff out accountable. It’s not good enough to just fire the people who put AI fluff into the world. We should be firing the people who should have been reading it, debating it, investigating it, and engaging it.
That we have AI experts discussing an AI policy without knowing that it’s been drafted by AI and has AI hallucinations is the problem, not the AI or how we use it.