So here’s the thing that’s been keeping me up at night, and it’s absolutely hilarious when you think about it: every time some executive in a $3,000 suit gets on stage and talks about “AI transformation” and “workforce optimization,” what they really mean is “we’re going to fire a bunch of workers and replace them with algorithms.”
What if they were describing their own job?
Remember Moravec’s paradox? The one that says AI is great at complex cognitive tasks but terrible at the “simple” stuff? Well, that might apply to the C-suite.
AI can play chess at superhuman levels. It can plan strategies, analyze markets, optimize supply chains, and even write those painfully boring quarterly reports that nobody reads anyway. You know what it can’t do? Make a decent pair of shoes.
Try getting a robot to craft a leather boot with the skill of a trained cobbler. Go ahead, I’ll wait. You’ll get something that looks like it was assembled by a drunk octopus. The stitching will be off, the leather will be mangled, and good luck getting it to adapt when it hits a knot in the material or needs to adjust for thickness variations. I know a thing or two about shoes. I buy a lot.
Same goes for a skilled electrician navigating the nightmare of wiring in a 100-year-old building. Also have been there. Or a plumber figuring out why your pipes are making that exact weird noise. Or a chef who can tell by touch when the dough is ready (my mum can do that by just watching it, but she plays in a different league). Or a carpenter who can see the grain in the wood and knows exactly where to cut.
Now let’s talk about what a CEO does. Strategy, right? Resource allocation. Market analysis. Risk assessment. Stakeholder management. Planning five moves ahead.
Sound familiar? It’s basically chess with PowerPoint.
And we know AI is good at chess. We know it’s good at:
Analyzing massive datasets to find patterns
Making predictions based on market trends
Optimizing resource allocation
Strategic planning
Risk modeling
Even generating persuasive communications, if you have been playing long enough with ChatGPT, lately
Didn’t we just say there are already AI tools that can analyze your company’s performance, suggest strategic pivots, and write the memo explaining it to the board? Is it true that an LLM could probably run your quarterly earnings call and nobody would notice the difference?
What I find really funny is that the same people warning us about AI displacement are the ones most easily displaced. The executive who goes on the news saying “We need to prepare workers for an AI-driven economy” is essentially saying “I’ve made myself obsolete and I’m too dense to realize it.”
Want to fire your CEO and replace them with Claude or ChatGPT? You’d probably save $10 million a year in salary and stock options. Want to replace your middle management layer with AI decision-making tools? There’s your next $40 million.
Meanwhile, try replacing the woman who’s been doing upholstery for 30 years. Try replacing the guy who can diagnose your car’s problem by listening to the engine for ten seconds (provided you drive real cars, not Tesla). Try replacing the artisan who hand-blows glass or the tailor who can fit a suit to your exact body.
You can’t. Not even close.
Why aren’t executives lining up to say “AI will replace management first”? Because that would require a level of self-awareness that’s incompatible with the executive mindset.
It’s way easier to look down the org chart than to look in the mirror.
The narrative is always “automation will transform low-skill jobs” (translation: we’re coming for the warehouse workers, the drivers, the cashiers). But the reality is that these jobs involve enormous amounts of sensorimotor skill, real-world problem-solving, and adaptive intelligence that we’ve barely scratched the surface of replicating.
Meanwhile, “high-skill” management jobs? They’re spreadsheets and meetings. They’re pattern matching and strategic thinking, exactly the stuff AI is demonstrably good at.
Ok Frag, so what?
Here’s my modest proposal: Let’s actually do this. Let’s replace the executives first.
Fire the entire C-suite. Replace them with AI decision-making systems. Keep the humans who actually make things, fix things, build things, and create things. The ones with calluses on their hands and skills in their fingers.
Take that $100 million you’re saving on executive compensation and distribute it to the workers who are actually keeping the lights on. The ones doing the work that AI can’t do. The ones with real expertise that took years of practice to develop.
Imagine that: A company run by AI management systems, but where the actual humans, you know, the craftspeople, the technicians, the makers, are paid well and respected for having skills that are genuinely irreplaceable. Because they are.
Here’s what we’ve learned from decades of AI research: The things humans find easy are extraordinarily hard for machines, and the things we find hard are often trivial for machines.
AI can beat you at Go or Chess. It cannot:
Tie a proper knot under tension
Adjust a recipe on the fly when an ingredient tastes off
Navigate a construction site without killing someone
Understand why the machine is making “That Weird Sound”
Cut someone’s hair and make them feel good about themselves
(Ok this one is really hard. Not even human barbers can do that)Fix the wonky toilet that’s been driving you crazy
Make a really good sandwich (and if you think this is easy, you’ve never been with an italian from the south)
So next time you hear some executive talking about “AI disruption” and “the future of work,” remember: they’re the low-hanging fruit. They are the ones playing checkers while thinking it’s 4D chess.
The real humans, the ones who build, fix, make, and create things in the physical world, they’re safe. Because Moravec’s paradox isn’t going anywhere.
And honestly? The idea of a world where AI handles the bureaucratic bullshit while humans focus on craft, creation, and tangible work? That actually sounds pretty good to me.
Just don’t expect the people currently in charge to volunteer themselves for obsolescence. That would require the kind of strategic thinking that, ironically, an AI would probably figure out in about 30 seconds or less.
P.S.
Before you get too excited about this revolution, let me tell you that this will never happen.
Why? Because the people who would need to make the decision to replace CEOs with AI are... the CEOs themselves. It’s the ultimate conflict of interest.
It’s like expecting politicians to vote themselves a pay cut. Sure, it might be better for society, more efficient, more fair. But you’re asking turkeys to vote for Thanksgiving.
The people with the power to implement AI replacement are exactly the people who benefit from NOT implementing it.
CEOs sit on each other’s boards. They golf together. They decide each other’s compensation packages. They control the narrative about “digital transformation” and get to aim it downward at everyone else. You think they’re going to turn that AI cannon on themselves?
I am sure the day they will get questioned about all this, they’ll commission studies showing that “leadership requires the human touch” and “strategic vision can’t be replicated by machines”, all while implementing AI systems to replace the workers who actually require human touch and irreplicable skills.
So yeah, firing all the executives is pure utopia. A beautiful thought experiment that will remain exactly that: a thought.
Now, I am done, for real. Till next time.
Did you know I am the host of podcast Data Science at Home? If you want more uncomfortable truths about AI, data science, and the tech industry delivered straight to your inbox, consider checking it out. We cut through the hype, call out the bullshit, and occasionally make you laugh while doing it.