AI, Jobs, and the Dinner Question I Wasn’t Expecting

This week, AI seemed to follow me everywhere. 

On Thursday, I was facilitating at EO Accelerator in Cape Town, teaching execution and legacy material that is increasingly being enhanced by AI. The prompts are better. The speed is incredible. Participants can move from blank page to usable output in minutes. Yet, despite all the technology in the room, I found myself repeating something I have said before.

“Stop talking to ChatGPT for a moment and talk to the smart people sitting next to you.” 

It sounds strange coming from someone who spends much of his week helping businesses implement AI systems, but the observation keeps repeating itself. Once people discover the power of the tool, they disappear into their screens. The room gets quieter. Heads drop. The conversation slows. The technology accelerates the work, but it can also accidentally remove the very thing that makes workshops valuable in the first place. The answer was, is, and will continue to be, in the room.

AI helps people think faster. Humans still help people think better. Then, over dinner this weekend, the conversation took an unexpected turn. A friend asked me what I charge companies to build AI solutions and automation systems. I answered him. He laughed and asked a question I had genuinely not considered in that way before. 

“How many jobs are you helping remove?”

It was not hostile. It was thoughtful. And it sat with me for the rest of the evening. Because while I understood the concern, it was not how I had been framing the work. My view has always been slightly different.

The businesses I work with are not implementing AI to reduce headcount. They are implementing it to achieve better systems, quicker responses, stronger outputs, and less administrative drag. Most of the work involves helping teams reclaim time from repetitive effort so they can spend more energy on higher-value activities. They want fewer client complaints, higher efficiency rates, and fewer mistakes.

I always make it clear that these solutions should be stress tested against suppliers and existing workflows. Not because replacement is the objective, but because the competitive environment is changing. If a supplier is not adopting available technology while their competitor is becoming faster, more efficient, and more professional through AI, then the market will eventually make the decision for them.

That is not really new.

Throughout history, we have had versions of this debate, whether it is in-source versus out-source models, industrialisation, automation, globalisation, or digitisation. The tools change. The pressure to adapt remains remarkably constant, and like that old Successories poster I had at my dad’s business, which depicted a huge wave with the slogan: “If you’re not riding the wave of change, you’ll find yourself beneath it.”

The difference now is speed. Caroline and I have started exploring a global tourism venture, and one of the first practical steps was to build a completely new website using AI. A task that would previously have required a subcontractor suddenly became something we could prototype ourselves, and that immediately raised another question: Would we still use a subcontractor? The answer is yes. Absolutely yes. But only if they were using the same technology and applying it to produce better work, more quickly, more cheaply, and on a greater scale.

The opportunity has not disappeared. The expectation has shifted. Not everyone has the time and the creative inclination to tinker with Claude and websites, and the person you hire to do it for you is going to create it faster, better and cheaper than ever before… and isn’t that a helluva democratisation of services in a country where there is such a big divide between haves and have-nots?

Every major technological leap arrives carrying two stories at once. One story focuses on what disappears. The other focuses on what becomes possible.

The Industrial Revolution removed roles while creating entirely new industries. The internet displaced jobs while creating careers that previously did not exist. AI feels similar, except the cycle is moving far faster than before. For a while, I found Elon Musk’s optimistic view compelling, the idea that technology would create abundance and free people for more meaningful work. Then you watch WALL-E again and remember the less inspiring version of the future where fat humanity ends up floating around while machines do everything and collective fitness becomes optional.

That is probably not where we want to land. And in South Africa, we have different challenges and opportunities.

We have enormous youth unemployment, while employers simultaneously complain that younger employees often arrive without the practical knowledge and skills required for modern work. Perhaps AI changes that equation, as Christo Wiese suggested in a recent chat with us. Perhaps it becomes an accelerant rather than a replacement. A tool that helps people become more articulate, more prepared, more informed, and more capable in less time.

If knowledge suddenly becomes cheaper and expertise becomes more accessible, that is not automatically bad news. In fact, access may be the bigger story. If websites become cheaper to build, content cheaper to create, research easier to perform, and knowledge easier to access, then more people participate. More entrepreneurs start. More small businesses compete. More people enter markets that previously felt out of reach.

And in a country as divided as ours between haves and have-nots, lowering the cost of access feels like something worth paying attention to. The question may not be whether AI replaces people. The question may be whether people use AI to become more valuable.

Start by looking at your own role and separating your work into three buckets: tasks that should be automated, tasks that should be accelerated, and parts that remain deeply human.

Then ask the same question of your suppliers, your team, and your business model. Because the opportunity may not be to remove people from the system. It may be to elevate them through it. 

ChatGPT Prompt:

Act as an AI transformation advisor. Review my role or business and categorise activities into three groups: “Automate”, “Accelerate”, and “Human Advantage”. Identify where AI creates efficiency, where humans remain essential, and what skills should be developed over the next 12 months to remain competitive. Present recommendations for me, my team, and my suppliers. 

Share on LinkedIn