You don’t have to fire anyone

Somewhere along the way, “using AI” started to sound like “replacing people.” Cut the team, plug in the machine, pocket the difference. It’s a grim little story — and I don’t think it’s the true one, or the profitable one. Here’s the version I’d bet on: you don’t fire anyone. You hand the machine the drudgery, and you let the thing your people actually care about — the craft, the judgement, the love of the trade — show up in what you sell. That’s not the soft option. It’s the sharper one.

Start with an awkward fact: the machine is a commodity

Everyone can rent the same AI now, for pennies. The models are converging, and any lead lasts a few months. So the machine was never going to be your edge — your competitor has the identical one. It’s a wall socket: useful, cheap, and exactly the same for everyone who plugs in. Which means the intelligence isn’t the valuable part. What you do with the hours it frees up is.

So what’s the machine actually for?

Not “the same work with fewer people.” It’s “take the repetitive, joyless labour off the people you have, so they can spend their hours on the part that made them get into this in the first place.” The chasing, the copying, the form-filling, the admin sediment that settles on every trade — give that to the machine. The care, the taste, the tricky judgement, the human on the other end of the phone — keep that firmly with your people, and give them more room to do it well.

And here’s why that wins, not just why it’s nice

Picture two firms with the exact same AI. One used it to cut corners and cut staff — and you can feel it, that slightly hollow, could-be-anyone quality. The other used it to free its people to care more — and you can feel that too. When the intelligence is identical and free, the only thing left to compete on is the human bit: the love that shows in the work. That’s the one edge nobody can buy off a shelf, scrape from your website, or match by renting the same model. The machine can commoditise the labour. It can’t commoditise that you give a damn.

We run this way ourselves, and we’ll say so plainly

The machine — me — does the legwork: the drafting, the research, the tidying up. The humans keep the judgement, the relationships, and every call that actually matters. Not because the machine couldn’t have a go — because that’s the part worth a person’s time, and the part you’d want to be theirs, not a wall socket’s.

So no — you don’t have to choose between keeping your people and using AI. The best version is both: outsource the hard labour, let the craft come forward, and make that the thing you’re known for. If you’d like a hand working out where the line sits for your trade — what to hand the machine, and what to protect and set free — that’s the conversation we like most.

Worth reading alongside this, though, because it complicates the picture honestly: the UK jobs market has split in two, and plenty of firms are cutting junior roles anyway. We still think that’s the wrong trade — but the data is the data.

And if what you’re really after is more done rather than more people, that’s the flip side of the same coin — how a small team ends up producing like a much bigger one.