I’m taking a break for a bit from AI social commentary and the increasingly tedious (and boring) work of watching the AI Confidence Bros and test tool vendors discover testing for the first time. I’m going to (try) to focus on some podcasts and AI integration e.g. what happens when these systems run into organizations, people, controls, incentives, legacy technology and reality.

I’m also developing some research under the working title, The Human-in-the-Loop Fallacy: Recursive Assurance in Agentic Software Development, looking at the increasingly absurd belief that HITL can provide meaningful oversight of AI systems at volumes and speeds no human can perform. So my next indictment of the AI-in-testing industrial complex might have to wait for my keynote at Agile Testing Days.

But before I take a break, I just want to add my two cents about posts I’m seeing more and more of like this one: “The AI job apocalypse is a myth. We need more human talent than ever before”. Sounds good, but entirely misses the point.

No one thinks humans disappear from in an AI economy. Capitalism has been remarkably inventive at finding things for people to do. The concern is about what kind of work remains, who controls it, how much it pays and who gets even richer off it.

As my friend Michael Bolton likes to paraphrase me, “show me the incentives, and I’ll show you the behavior”.

And I think it’s increasingly clear that the goal of Big AI, judged by its incentives (and actions), is not to eliminate every worker. It’s to make labor cheaper, weaker and more replaceable with a smaller group of extremely wealthy owners sitting above a permanent underclass of supervisors, labelers, and exception handlers for systems they neither understand nor control.

And almost as if on cue comes the latest “doom troll” (ala Ed Zitron) designed for maximum scare value in the form of We Must Act Now. Four sentences signed by more than 200 economists, researchers and technology executives warning that AI could transform the economy faster than the Industrial Revolution and cause large-scale job displacement.

We’re going to see a lot more of this crap as the AI bubble continues to strain under its mounting costs and debts coupled with the growing uncertainty over whether any of this stuff is going to be worth it. When you’re up against it, the inevitability of “too big to fail” becomes your only asset.

And what really sucks is it doesn’t have to be like this. In Building Pro-Worker Artificial Intelligence, the authors explain that AI COULD be designed to make human expertise more valuable, create new tasks and extend worker judgement. But that version is systematically underfunded. Companies favor automation, deskilling and “expertise-leveling” because those approaches transfer more of the power and wealth from workers to managers and shareholders.

The British public seem to get it. The UK AI Compass surveyed nearly 3,000 people and 69% were concerned about AI’s direction, 85% wanted stronger government action, 68% believed the gains would go to billionaires and corporations, and only 10% thought ordinary people would benefit. Seventy-four percent expected AI to reduce jobs; just 10% expected it to create them.

The system being built does not work for us and people know it. People want rules, accountability and AI that works for them, not imposed upon them by a handful of companies whose definition of social progress begins and ends with shareholder value.

So that’s what pisses me off about these naive and overly optimistic articles opining that “we need more human talent” while companies spend billions trying to make talent cheaper, pay people less, make them more interchangeable and leveraged to the eyeballs. So the coming AI job apocalypse is not a myth, but the people carrying water for Big AI lack the imagination to understand the result will be more change than destruction.

AI has entered a new phase of public acceptance moving from curiosity, to hype, to now outright hostility. What was once largely an online debate has begun spilling into the physical world, with threats against executives, data centers, and employees becoming a genuine security concern. Let’s hope we change course before we find out why billionaires are building those have built those bunkers.