Almost every discovery call I have right now opens with the same quiet worry, usually phrased as a half-joke: “So… is this the bit where you tell me to make people redundant?”
The honest answer is no. And I want to be clear about why no, because the way most people are being sold AI deserves the scepticism.
The fear is reasonable. The framing is wrong.
The hype machine has spent two years telling small businesses and not-for-profits that AI is a headcount-reduction tool. If that were the goal, I’d understand the dread — most of the teams I work with are already stretched thin. An overloaded NFP doesn’t have “spare” people to remove. They have one person doing IT, comms and half the admin, and they’re drowning.
So when I look at a team like that, I’m not asking “who can AI replace?” I’m asking a completely different question: what is the draining, repetitive work that’s stopping these people from doing the part only humans can do?
AI should take the work nobody wants, so your people can do the work only they can do.
That’s the whole philosophy. It’s not a slogan I reverse-engineered for a website — it’s what 15 years of looking after Australian businesses taught me. The value was never in the busywork. It was always in the judgement, the relationships, and the urgent calls that genuinely need a person.
What “complementing” actually looks like
Here’s a real shape of it, anonymised. A not-for-profit was losing around 500 phone calls a month. Stretched staff, burnt out, and the people who genuinely needed urgent help couldn’t get through.
The wrong “AI” answer is a chatbot that replaces the front desk and frustrates everyone. What we actually did: read the phone-system logs, found why calls were dropping, fixed it — and then set up an assistant to qualify incoming calls and route the routine ones to self-serve answers. The urgent lines stayed free for the people who needed a human.
Nobody lost a job. The team got their time back for the interventions that actually need them. That’s the difference between AI at your people and AI for your people.
The test I’d apply before spending a dollar
If you’re weighing this up for your own team, here’s the question I’d put to any vendor — including me:
- Does it keep a human in the loop on anything that matters? If not, walk away.
- Does it start small and prove value, or does it want a big platform commitment up front? Small wins first, always.
- At the end, is your team more capable, or more dependent on the vendor? The right answer leaves you in control.
That last one is the real tell. A lot of “AI solutions” are designed to make you need the vendor forever. The work I’m proud of does the opposite — we train the team to run the thing themselves.
So, will AI replace your staff?
Not if you do it properly. Done right, it does the opposite: it protects your people from burnout and frees them up to do their best work. Done wrong, it’s an expensive way to annoy your customers and hollow out your team.
I know which one I’d rather build. If you want to figure out the one task AI should genuinely take off your plate — and be told honestly if the answer is “none yet” — that’s exactly what a discovery call is for.
