AI doesn't innovate. People do.

We've seen this play out before with other technologies. Blockchain, the metaverse, IoT. Each wave brought genuinely useful innovation, wrapped in a layer of inflated expectations. What makes AI different is that the core technology truly delivers. Language models, computer vision, and generative tools are producing real results in real organisations.

The challenge is that the conversation around AI often follows the same old script: start with the technology, then figure out what to do with it. And in our experience, that order tends to produce disappointing outcomes.

GPT, Claude, Gemini, open-source alternatives. These models are available to pretty much anyone with an internet connection. Which means the technology itself isn't what sets organisations apart. What makes the difference is how you decide to use it.

The AI projects that tend to work well are the ones that begin with a genuine problem. Something like: where are our customers getting stuck? What's taking our team way too long? Which decisions are we making on gut feeling when we could be using data? When you start there, AI often turns out to be a powerful part of the answer. But it works best as an answer to a well-defined question, not as a starting point.

We see it happen regularly. A company launches a chatbot because a competitor did, without really checking whether customers would rather just have a clearer FAQ page. An internal AI tool gets built to "boost productivity", and six months later barely anyone is using it because it's solving a problem nobody actually had. A recommendation engine surfaces technically relevant content that somehow still misses the mark in context.

These aren't technology failures. The models work fine. The issue is usually that not enough time was spent understanding the actual problem, talking to users, or defining what success looks like.

This might sound obvious, but it's worth saying: innovation has always been about people. About noticing something that could be better, imagining an alternative, and finding a way to make it happen. AI is incredibly good at accelerating parts of that process. It can crunch through data, generate options, spot patterns, and handle repetitive tasks so people can focus on the work that actually requires judgment and creativity.

But it can't tell you which problem matters most. It can't sit in a room with your customers and understand what frustrates them. That part is still very much a human skill, and honestly, we think that's a good thing.