The leaders who get AI right are the ones who do the job first

It’s easy for leaders to find themselves running a business from behind a spreadsheet. They see the dashboards, the KPIs, the quarterly numbers, and from that vantage point, everything looks manageable. For a while, it is. But eventually, the gap between reality and spreadsheets catches up with them: either the information stops flowing upward the way it used to, or the team ends up operating in a completely different reality from the one the leader believes they’re running. The culture the leader thinks they’ve built and the culture that actually exists on the ground quietly drift apart.

 

This matters more than ever as AI moves from the periphery of business into the centre of how work actually gets done. Because a leader who is disconnected from the work doesn’t just make worse decisions - they set a precedent. If the person at the top isn’t close to the coalface, why would anyone beneath them be? Delegation has a habit of sliding downward: the most important, most judgement-heavy tasks get pushed further and further down the chain, to the most junior person available, or increasingly, to the algorithm. Left unchecked, that’s how you end up with your least experienced resource - human or machine, of course - quietly making your most consequential decisions.

 

The single best predictor of whether an AI or technology rollout succeeds isn’t the sophistication of the model or the size of the budget behind it. It’s whether the people designing it have actually done the job it’s meant to support - not shadowed it, not surveyed it, but done it.

 

The question every business leader deploying AI should be asking themselves is simple, but uncomfortable: is this technology making people better at their jobs, or is it quietly replacing them without improving the quality of the work? One way to tell is to look at what happens to the time it frees up. If automating a task simply clears the way for another task that can also be automated, you haven’t empowered anyone – you have just replaced people with technology. But if that freed-up time goes toward the parts of the job that genuinely require a human - – judgement, care, conversation, relationships – then technology has done exactly what it should.

 

I came to understand this properly in social care, an industry I didn’t come from and had to learn before I trusted my own instincts about it. Early on, I was reliant on trade press and second-hand reports from the team, but as a leader I needed to feel confident in the decisions I was making. So, I went and lived in a care home for an entire week, not as a visitor but as a resident. I remember talking to one of the managers when she suddenly had to excuse herself. She’d promised a resident they would sit down together over a cup of tea, and she would paint her nails – and she didn’t want to keep her waiting. It sounds like a small thing, but that was precisely the point: the value of technology is found in the moments that have nothing to do with technology at all.

 

None of this is unique to care. Every sector facing AI adoption right now is wrestling with the same fork in the road, whether it is retail staff being freed from admin to spend more time with customers, or recruitment teams spending less time screening CVs and more time actually speaking to candidates. Across industries, the underlying question doesn’t change: is the freed-up time being reinvested in what people do best, or simply being used to squeeze out more of what a machine could do anyway? Get that wrong at scale, and you don’t just end up with disappointing technology - you end up with a workforce that has been quietly deskilled and demoralised, often without anyone consciously deciding that should happen.

 

Confidence in that judgement, I’d argue, is not something you can get from a dashboard. It comes from proximity to the work itself. Leaders who stay close enough to see how time is actually being spent - not how it’s meant to be spent, according to the org chart - are the ones who catch the drift before it becomes a crisis.

 

So before signing off on the next AI investment, the question worth asking isn’t only "what will this automate?" It’s "what will my people do with the time this gives back?" If you don’t already know the answer, that’s usually a sign it’s worth going and finding out yourself, on the ground, before the decision gets made for you.

 

---

 

Kehan Zhou is Group CEO at Person Centred Software

 

Main image courtesy of iStockPhoto.com and DragonImages