When ChatGPT first came out, I was hooked. I used it for everything. I asked it how to remove a stain from my favourite shirt, I used it to edit presentations, and I leaned on it to help me draft tricky emails. It became my daily co-pilot. I started to wonder, if this makes my life easier, what could it do for my team?
So I brought it into the workplace. I built AI-powered workflows for tasks like generating accessible alt text for images and helping people make decisions faster. I believed I was giving my team a productivity superpower.
The reality was humbling. Almost no one used them.
As operations leaders, we often see efficiency as an obvious win. I thrive on making work simpler and more streamlined. However, I learned the hard way that a process is only as valuable as the number of people who actually use it.
This seems obvious now. I would see someone drowning in emails and think, I know exactly how to fix that. I would then spend hours building a clever AI-enabled workflow. The result was usually a polite message asking if we could turn it off.
Lesson learned. Even the most elegant process will fail if it is not solving the problem your team feels most urgently.
Here is a controversial opinion. People have to discover their own way in with AI before they will embrace it.
I assumed my team would use AI exactly the way I did. They did not. Unless they have the chance to explore on their own terms, choose their preferred tools, experiment with prompts, and figure out what is useful, adoption will stall.
As a leader, your role is not to impose your favourite AI workflows. It is to set the tone, provide access, and create a safe space for experimentation.
Half measures do not work. I read about how Zapier rolled out AI across their company. What stood out was that every employee reimagined their workflows from scratch. It was not about adding AI onto existing processes. It was about redesigning with AI at the core.
To make AI stick, you need this kind of cultural shift. If only a few people are using it while the rest work the old way, the momentum dies.
AI can do it. The real question is whether your people want it to. The technology is not the barrier. The barriers are change, trust, and whether it feels relevant to the day-to-day work of your team.
Next time, I will start with conversations, not automations. I will let my team play, experiment, fail, and find their own wins. Only then can AI become part of our culture instead of another unused tool gathering dust.
