We have spent decades and billions building machines to move humans around. Tanks, fighter jets, ships. Huge engineering effort, all of it organised around carrying people into a fight.
Then look at Ukraine. Out of necessity they built their strategy around something else entirely. Cheap drones, smart targeting, machines that are expendable because there is nobody inside them. A drone costing a few thousand dollars can threaten equipment worth orders of magnitude more.
That isn’t a better tool bolted onto the old way of fighting. It’s a different answer to what the whole effort is optimising for. The old model optimised for moving people. The new one optimises the whole strategy with AI sitting at the core.
That’s the shift. Optimise with AI, not for AI. It’s the same shift every business is facing, just with lower stakes.
Which is why the usual question, are you using AI yet, is the wrong one. AI is an optimisation engine. Point it at something and it drives hard toward it. What matters is what you’ve pointed it at, and how it gets there.
Most businesses haven’t decided
A lot of businesses bring AI in without ever deciding what it should be optimising for.
So everyone decides for themselves. One person uses it for speed. Another for tidy code. Another just to get through a review without really understanding the work. The AI helps all of them, because that’s what it does. It has no opinion about whether any of it serves the business.
You end up with output that technically works but isn’t anyone’s vision of anything. Everyone feels productive. Nothing pulls the same way.
Ukraine’s edge wasn’t just the drones. It was that the drones served one strategy. When everyone points AI in their own direction, you lose exactly that.
The how matters as much as the what
It’s not only what you point AI at. It’s how you want it to get there.
A business can optimise for growth by looking after its customers, or by squeezing them. Same goal, completely different how. AI will run either one. It won’t pick the version that feels like you. That part is still on us.
This is where a person in the loop earns their place. Not as a handbrake, but as the one asking does this feel like us, does this serve what we’re building. AI optimises. It doesn’t have taste.
AI won’t fix the model
Point a powerful optimiser at a goal you haven’t really thought through and it just chases that goal faster and at a bigger scale. AI doesn’t fix a bad business model or an unclear direction. It runs whatever you give it, harder.
The businesses struggling with AI mostly aren’t being let down by the technology. They handed an optimisation engine to an organisation that hadn’t decided what it was optimising for.
The question worth sitting with
Before how do we use AI, there’s a better one.
What should our AI be optimising for, and how do we want it to get there?
Get that clear and the rest gets simpler. Leave it open and you just get more of everything, faster, pointed in every direction at once.
So, what is your AI optimising for right now? And did anyone actually decide that?
Nigel Price is the founder of Digital Discovery Group, specialising in ecommerce strategy, digital transformation, AI-powered platforms, and managed cybersecurity services for small and medium businesses.
Comments