How can AI actually help my business?

Let me answer this one from the inside.

AI has completely changed how I work. Not by turning my practice into an AI business, not by creating some specific AI product, but by making me a faster, sharper version of myself. I do the same work I’ve always done (thinking through digital problems, designing solutions, testing ideas) but at a speed that would have seemed impossible five years ago.

That’s the honest version of how AI has helped my business. Not transformation. Acceleration.

Where AI helps most: prototyping

The clearest example I can give is prototyping.

Over 25 years, I’ve watched the tools evolve. Grayscale wireframes in Corel Draw. SketchUp models. Figma designs. High-fidelity interactive prototypes in Axure. Each iteration got closer to the real thing: a way of showing a client or a user what something would feel like before committing to building it.

Today, my prototypes are living code. Not production-ready, not something that could go live, but real-world working interfaces that users can actually experience rather than imagine. The prototype used to represent the thing. Now it is the thing: close enough that you get genuine, useful reactions from real people, not polite nods at a static mockup.

AI made that possible. What used to take weeks now takes days. What used to take days now takes hours. The ability to test whether you’re building the right thing, before you commit to building it, has never been faster or cheaper.

The real value comes from using AI constantly

But here’s what I’ve noticed about where the real value kicks in.

It’s not when you use AI occasionally. It’s when you’re using it constantly: when it becomes part of how you think through every problem, not a tool you pick up for specific tasks. That’s when the compounding starts. That’s when you find yourself using the thing you’re building while you’re building it, which creates a feedback loop that used to take months and now happens in real time.

I’m aware this is a specific situation. I work alone, or in small focused teams. I have 25 years of judgment about what questions to ask and what answers to distrust. I can feel when something’s wrong even when AI is telling me it’s fine.

I watch what’s happening in larger organisations with long eyes.

What worries me in larger organisations

What I see concerns me.

In bigger organisations, AI seems to be making more people think less, not more. The pattern is consistent and, I’d say, endemic at this point: someone pops a question into an AI tool, gets an answer, and presents that answer, without actually thinking through the problem themselves first. The thinking step gets skipped. The AI becomes a bypass, not an accelerator.

This is the same problem we’ve talked about throughout this series, just in a new form. The shortcut feels productive. The output looks credible. But if the question was wrong, or the framing was off, or the problem wasn’t properly understood before the AI was consulted: the output is just a well-written wrong answer. Efficient mistakes are still mistakes.

The question that matters more than how to use AI

The pace of change makes all of this harder to navigate.

From my first serious use of AI to the frontier models available today is a genuinely mind-boggling distance. The tools I was using two years ago feel almost quaint. What’s available now, and what’s coming, makes the question “how can AI help my business?” almost impossible to answer completely, because the answer is different every six months.

What I can say is this: AI will replace some jobs. Not all of them, not now, and probably not in the way most people fear. But some of what people do today will simply not be valuable tomorrow. And that raises a question I find more useful than “how do I use AI?”

What is my actual value?

The businesses and individuals who sit honestly with that question, who identify which parts of what they do aren’t valuable anymore and accept it, are the ones who can use AI to amplify what remains. Not as a replacement for thinking. As a multiplier of the right thinking.

If every person in an organisation went through that process genuinely (understanding their true value, letting go of the parts that AI has made redundant, and using AI to go further with what’s left), something very interesting could happen at scale. The potential is real.

But it requires honesty about what AI is actually doing versus what people wish it were doing. And it requires someone willing to ask the hard questions before reaching for the tool.

That part hasn’t changed. It never does.

Garth Shoebridge has been using AI as a core part of his digital practice for several years: building with it, thinking with it, and watching what it does and doesn’t do in real business contexts. If you’re trying to work out where AI can actually help your business, start with a conversation.