They said AI would take half the jobs. It didn't.

Predictions vs Reality

Even if AI doesn’t evolve any further, it has already transformed my work. Everyone predicted the opposite. In an article I read today by Cal Newport, the loud predictions from the AI world were laid out: half of all jobs would disappear, half of entry‑level white‑collar jobs would be automated. Dario from Anthropic said that, then later changed his stance to say only parts of jobs would be automated. Mustafa from Microsoft AI claimed human‑level performance on most tasks within 12 to 18 months, yet we still don’t see that. Sam Altman warned of a jobs apocalypse and now says he’s delighted to have been wrong. All of those predictions and walk‑backs are documented; I’m not inventing them.

Marketing Noise

Since I started studying AI, I’ve felt that the loudest predictions were marketing. They were exaggerated, biased, and designed to sell. The real discourse becomes more realistic as AI improves.

Small‑Scale Tooling

What I’m seeing now is owners hacking together little tools. Small business owners are building throwaway solutions, simple spreadsheets, login pages, databases, that solve a specific need but don’t replace senior engineers. Those tools are not the digital equivalent of a power loom that makes large numbers of jobs superfluous.

Delegation Over Building

My clients often delegate because they have no time. Even when AI makes things faster, they still lack the bandwidth to build and fine‑tune prompts. They prefer to pay someone to solve the problem rather than spend time fixing the AI.

Last week I went to my brother’s jiu‑jitsu gym. He was selling a subscription system he’d built, and I helped pull data from the gym’s turnstile. The owners didn’t even know that integration was possible; the data was there but locked away. AI did the typing, but the judgment stayed with me. AI wrote the bridge script for my brother’s system, and I identified the gap no one had perceived, made the security decision, and launched it with real data. The human decision‑making part remained essential.

My Own Toolkit

Since February I’ve turned my routine into 49 repeatable skills and more than 300 scripts that handle tasks without wasting AI tokens or generating new code. These small tools are exactly what owners do: they’re layered one after another to get the job done. I also build larger, more complex systems when a single solution can solve a big problem more efficiently.

Raising the Bar

AI isn’t replacing people like me; it’s creating more demand for productive humans. Clients now expect me to be more productive, to juggle three or four tasks at once, and to be proactive. The parts of work that don’t move, understanding client gaps, providing solutions, maintaining relationships, cannot be automated. AI took the typing, but the actual work, especially the judgment and creativity, remains human.

Vocation

I believe in the idea of a calling, a vocation. The Catholic notion of vocation is that each person has skills given by God that incline them toward certain work. For entrepreneurs, the calling is to grow a business, employ people, and have community impact. That’s why we delegate technical tasks to developers, whose calling is technology. I work with AI not just for money or hype; I enjoy it and see it as a way to serve my clients.

That’s it for today’s video. See you in the next one.