At a recent CTO roundtable hosted by MCS Group in Dublin, one theme echoed across the room: doing more with less. It sounds practical—even aspirational. But as the conversation deepened, it became clear that this phrase risks becoming a euphemism for burnout, disillusionment, and missed opportunity if not unpacked properly.
In today’s market, AI is no longer optional. It’s assumed. If you want more headcount, you’d better first demonstrate why AI can’t solve your problem. And to some extent, that makes sense. Leaner teams, tighter margins, greater demands—every lever must be pulled. But somewhere along the way, we’ve started confusing efficiency with productivity, and the fallout is becoming increasingly visible.
(Strictly speaking, efficiency is doing things right; effectiveness—or efficacy—is doing the right things. But in many tech conversations, those lines blur—and that’s where risk creeps in.)
The Efficiency Illusion
Efficiency is about reducing waste. Productivity is about increasing output. You can be efficient but unproductive if what you’re doing isn’t valuable. And you can be productive but inefficient if your systems are poorly designed.
AI tools like GitHub Copilot have certainly boosted code velocity for senior engineers. But across teams, the results are uneven. Juniors often don’t understand how or when to use these tools effectively. Some fear them. Others misuse them. The result? Increased tension, inconsistent quality, and diminished mentoring opportunities.
Worse, some organisations have used AI as a justification to cut support, QA, and mid-level roles. Yet over half of those leaders now regret it. Efficiency should not come at the cost of resilience.
The assumption that AI will unlock productivity overnight is proving naive. From GitHub Copilot to Make.com, tooling is abundant—but unless teams are data-ready, properly trained, and culturally aligned, the result is uneven output, rising tension, and even attrition.
Teams Are Systems Too
There’s also a less discussed but equally important dimension: the human cost of “doing more with less.” When we freeze hiring, we don’t just trim budgets. We stop the flow of new energy, diverse thinking, and constructive friction. New graduates and junior engineers bring fresh perspectives—not just more hands on keyboards.
Without that renewal, teams calcify. We lose dynamism. Innovation withers. As a CTO, I’ve seen this first-hand: the best ideas often come from the least expected voices—if they’re in the room.
The Data Bottleneck
Another unspoken truth: many teams can’t “just use AI” because their data infrastructure isn’t ready. Poor data lineage, inconsistent semantics, and silos are where efficiency goes to die. The model isn’t the problem. The inputs are. And unless architectural debt is addressed, even the best AI will amplify the wrong signals.
Flattery and False Confidence
One point I made at the roundtable triggered a few laughs—but also some nods. I compared AI to a seductive figure at the bar: flattering, attentive, always telling you what you want to hear. That’s by design. These tools are built to placate. But in business, especially in high-stakes decision-making, what we don’t want to hear is often what matters most.
As leaders, we need to interrogate the outputs. Ask: what shouldn’t we do? What’s missing? What’s the second-best answer? Otherwise, we risk building castles on sand—beautiful, logical, convincing sand.
Towards a Smarter Definition of Efficiency
Efficiency isn’t about fewer people. It’s about fewer bottlenecks. It’s not about doing the same work faster—it’s about redesigning the work altogether. And that requires architectural thinking, data clarity, and, perhaps most importantly, psychological safety within teams to challenge assumptions and propose alternatives.
What doing more with less should mean:
- Less redundancy, more system leverage.
- Less shallow automation, more domain context.
- Less fear-driven adoption, more iterative, supported experimentation.
Close: Partnering with the Right People
That’s why events like MCS’s roundtable matter. They curated a discussion that avoided the usual cheerleading and focused instead on what’s actually happening inside real teams—the tension, the experiments, the contradictions. We need a space not just for sharing challenges, but for reframing them. If we want to do more with less, we have to start by asking: more of what? and less of what? Until we get those answers right, AI is just another voice telling us we’re doing great—whether we are or not.