After years in digital transformation—across tech shifts, organisational overhauls, and a few “next big things”, you develop a bit of a radar for hype.
Right now, we’re at peak AI buzz. Everywhere you look, there’s a new tool, a new use case, a bold claim. And yes, some of it is real. The tech is moving fast, and it’s impressive.
But I’ve been here before. We’ve seen similar waves with automation, cloud, big data, you name it. The pattern is familiar: big promises up front, real complexity underneath, and eventually, the realisation that transformation takes more than technology.
What actually drives impact?
Not the flashiest pilot. Not the biggest vendor.
It’s the projects that:
- Solve an actual business pain
- Fit into real workflows
- Make someone’s job easier, faster, or more accurate
- Respect the constraints of the business
I’ve worked on transformations that succeeded quietly and failed loudly. The difference was rarely the tech. It was how grounded the effort was. Did we involve the right people early? Did we measure real outcomes? Did we listen when things didn’t land?
AI isn’t magic. It’s just new.
Rodney Brooks wrote back in the ’80s, “There is no magic in AI.” He was right then, and he’s still right.
It’s tempting to treat AI as a shortcut, a way to leapfrog inefficiencies or reinvent business models overnight. But the truth is, AI works when it’s pointed at clear problems, with realistic goals, and strong governance behind it.
Transformation happens through small, thoughtful wins, each one building trust. Trust from users. Trust from leadership. Trust from the people you’re asking to change how they work.
A few things I’ve learned (sometimes the hard way):
- Start with the problem, not the tech. If it’s not solving something that matters, don’t do it.
- Don’t scale what doesn’t work. A cool demo is not the same as a sustainable solution.
- Bring people along. No one adopts what they don’t understand or weren’t part of shaping.
- Be honest. About the limitations. About the risks. About what success actually looks like.
- Measure what matters. Not just “we deployed the tool,” but “this saved X hours” or “reduced turnaround by Y%.”
Most importantly: Trust is the real infrastructure.
If people don’t trust the tools, the insights, or the intent, you won’t get adoption, no matter how good the model is. AI success is built on human trust, not technical power.
So if you’re leading an AI initiative: stay curious, stay humble, and stay close to the people whose work you’re trying to change.
That’s how real transformation happens.