The AI Layoff Trap: Why Over-Automation Hurts Everyone

Lachlan Reed

[warmly] Welcome to the show — I’m Lachlan Reed here with Simon Carver, and Simon, I want to start with a number that should make every executive sit up straighter: in 2025 alone, more than 100,000 tech workers were laid off, and in more than half those cases AI was cited as a primary driver.

Simon Carver

[concerned] More than HALF? That’s the part I can’t shake. Because once you say “AI did it,” you’re not talking about one bad quarter anymore — you’re talking about a logic spreading from company to company.

Lachlan Reed

Exactly. And the paper on this — The AI Layoff Trap, March 2026 — says the ugly bit out loud. Firms can SEE the cliff. They know laid-off workers are also customers. Lose enough wages, and you hollow out demand. But each company only feels a slice of the damage, so each one still races to automate. Bit like everyone hearing the thunder but still camping under the same gum tree.

Simon Carver

[questioning tone] Wait — “only feels a slice” is the key phrase there. So if my company automates support or ops, I get the savings, but the pain of those lost paychecks gets spread across the whole economy?

Lachlan Reed

That’s it. The firm keeps the cost savings from cutting labour, but the demand loss gets smeared across rivals too. The paper models it as a proper arms race. In the frictionless version — and this is the memorable bit — it turns into a Prisoner’s Dilemma. Every firm automates because it’s rational individually, even though they’d all be better off if they restrained themselves together.

Simon Carver

[reflective] So this isn’t just “greedy bosses being greedy.” It’s worse, actually. It’s decent-seeming managers inside a system that rewards the wrong move. If they pause while rivals don’t, they look slow. If they cut, they hurt the market they need. It’s like everyone’s sawing off a branch while arguing about who gets the bigger half.

Lachlan Reed

[laughs lightly] Dark, but yes. And the paper says more competition makes the problem WORSE, not better. Better AI makes it worse too. Wage cuts don’t fix it. Free entry doesn’t fix it. Even UBI, worker equity, upskilling — helpful in pieces, but they don’t stop the basic incentive to over-automate.

Simon Carver

You know what really lands for me? “Over-automate.” Not automate — over-automate. That means the problem isn’t the tool. It’s when firms use AI like a layoff machine instead of a redesign tool. And then the damage isn’t just unemployment. It’s weaker consumers, shakier trust, and workplaces that feel cold enough to make people hide rather than contribute.

Lachlan Reed

[serious] Spot on. Once people think the company sees them as a temporary cost, not a future capability, the whole culture goes crook. Middle layers get hollowed out, leadership pipelines dry up, and you end up with what some folks are calling a barbell company — a few strategists up top, a giant automated engine down below, and not much human connective tissue in the middle.

Simon Carver

And that connective tissue matters in a crisis. It’s the person who knows how the client actually reacts, the manager who spots fear before it becomes attrition, the team lead who can say, “Hang on, the dashboard says one thing but the humans are telling us another.” Lose enough of that and efficiency starts looking a lot like fragility.

Lachlan Reed

[reflective] Yeah. That’s the trap. Layoffs can look brilliant in the spreadsheet and rotten in the real world. The market doesn’t just need productive firms. It needs paid people with enough confidence to buy stuff, enough trust to stay engaged, and enough humanity left in the workplace to keep the whole show from wobbling over.

Simon Carver

[gently] Which takes us to the more hopeful fork in the road. The better books on this don’t say, “Pretend AI isn’t coming.” They say, “Stop using it to erase meaning.” There’s that great reframing from The Last Job You’ll Ever Hate: this is not the end of work — it’s the end of meaningless work.

Lachlan Reed

And that’s where Maya comes in, right? Before AI, her day’s just a dog’s breakfast — inbox avalanche, status decks, data copied across three systems, meetings about meetings. After AI, the admin grunt gets stripped away. The reports build themselves, the updates get summarised, the nonsense shrinks, and she finally does the bits only she can do: judgment, coaching, sorting sticky trade-offs.

Simon Carver

Yes — Maya is memorable because the change isn’t abstract. It’s not “productivity rose 17%.” It’s that she gets 90 minutes of protected thinking time. She can mentor a junior colleague in the afternoon because she isn’t buried in slide formatting. Her work stops feeling like digital traffic control and starts feeling like leadership.

Lachlan Reed

[curious] So let me try to say it back. Good AI removes busywork, not dignity. It takes the repetitive muck off the plate so the human job can move UP the ladder — more creativity, more empathy, more problem-solving, more real responsibility. Have I mangled that?

Simon Carver

[warmly] No, that’s beautifully put. And it lines up with the other source too — Club Genius keeps hammering this idea that “people are the solution,” and that whole warning about “management by spreadsheet.” If technology just turns everybody into abstractions, employee IDs, metrics, counters — then you haven’t modernized work, you’ve dehumanized it.

Lachlan Reed

That line sticks. People are the solution. Not because software’s useless — software’s grouse when it kills admin sludge — but because the actual value in the next economy comes from things machines still can’t carry on their back: trust, ethics, taste, ambiguity, giving a nervous customer confidence, leading a team through a messy call.

Simon Carver

And there’s a practical leadership lesson here. If AI removes the junior tasks and the coordinating tasks and the repetitive tasks, then companies can’t just say, “Great, fewer humans.” They have to redesign roles around what only humans can do. That means training for discernment, conflict resolution, storytelling, mentorship, ethical oversight. Not just tool use — human use.

Lachlan Reed

[laughs] Human use — that’s a funny phrase, but fair dinkum. Because if you automate for efficiency alone, you get a brittle outfit: lower headcount, weaker trust, no bench of future leaders. But if you automate to elevate people, you get a stronger workforce, not a smaller soul.

Simon Carver

[softly] And that’s the choice, really. One future treats AI like a clever cost-cutter. The other treats it like a chance to restore dignity — to let nurses nurse, managers mentor, analysts interpret, teachers teach. One version hollows people out. The other makes them more capable, more relevant, more alive in their work.

Lachlan Reed

So maybe the question for leaders isn’t, “How many jobs can we remove?” It’s, “What human work are we finally free to take seriously?” [short pause] That’s where we’ll leave it.

Simon Carver

[warmly] Thanks for spending a few minutes with us on The Human Workforce.

Lachlan Reed

If this one got your gears turning, share it with someone wrestling with AI at work. Catch you next time.