The AI redundancy fallacy
Why workforce cuts sold as "cost optimisation" don't add up. SMEs are already the skeleton. AI should give them room to breathe, not excuse thinning it further.
Article body
The claim
Headlines tell a simple story: AI replaces workers. Companies announce redundancies and frame them as cost optimisation. "We're adopting AI to streamline operations." The implication: fewer people, lower costs, more efficiency.
It sounds logical. It's also a fallacy.
The reality: costs add up
AI doesn't run for free. Every agent, every API call, every token consumed has a price. Cloud AI scales with usage: more volume, higher bills. Self-hosted AI needs infrastructure, power, and maintenance. Orchestration tools, integrations, support. It all mounts up.
Cut your workforce and you might save on salaries. But you're trading one cost for another. And you're losing something that doesn't show up on a P&L: the people who know how things actually work.
SMEs are the skeleton
Subject Matter Experts are the skeleton of your organisation: the people who've been in the business for years, who know the quirks, the exceptions, the "why we do it this way". They hold the institutional knowledge. They spot when something's wrong. They have the human instinct no model has.
When you cut SMEs to "optimise," you're not just losing headcount. You're eroding the structure that keeps the business standing. Processes break. Quality slips. Nobody knows why that customer always gets special handling. The skeleton gets weaker.
Adopting AI isn't about reducing workforce. It's about streamlining process and giving SMEs room to breathe.
What "room to breathe" means
AI can take the repetitive work off your experts' desks. Triage tickets. Draft responses. Summarise documents. Cross-reference systems. The stuff that burns hours without adding much value. Free your SMEs from that, and they can focus on what actually matters: judgment, relationships, edge cases, the work only a human can do well.
- Streamline, don't replace: AI handles volume; humans handle nuance.
- Preserve the skeleton: keep your SMEs. Give them better tools, not fewer colleagues.
- Measure the right thing: ROI isn't headcount reduction. It's time saved, quality improved, and experts freed to do expert work.
A better calculus
If your AI strategy starts with "how many people can we cut," you're asking the wrong question. The right question: "how do we give our best people more leverage?"
That means process design. It means knowing where AI adds value and where it doesn't. It means measuring cost and ROI honestly, including the cost of losing the people who hold your business together.