Every time a new technology hits the market, we find ourselves coming to the same conclusion: Software is always less interesting than the people who use it.

And AI’s no exception.

In this series, we’ll be diving into the AI-empowered people at CreateFuture – asking them how, when and where they use AI, and exploring their experience of AI adoption.

Today we spoke to Angus Allan, Senior Product Manager at CreateFuture.

A former tech company founder from New Zealand, Angus is one of CreateFuture’s AI experts – a role which has seen him quoted in publications ranging from Wired to LeadDev and ITPro.

We sat him down to explore his view on AI at CreateFuture – how we approach it, how it’s evolving and how other organisations can help make their AI adoption successful.

How would you describe CreateFuture’s approach to using AI in-house? What’s your philosophy on it?

We spent a long time trying to figure out our philosophy, and we eventually boiled it down to one headline:

“Creating Tomorrow, Together, Today: Powered by people, accelerated by AI”  

In practice, that means that AI on its own can’t make us more effective or efficient. It can’t replace people. In fact, people are absolutely crucial if you want your AI implementation to make a difference.

And that means that people need to sit at the centre of any AI strategy. Your culture, your training and L&D, and your people’s mindsets all need to be prepared to embrace AI.

And AI needs to, first and foremost, serve your people – not replace them. We’re not just interested in building AI-powered products. We want our clients – and our own people – to be able to explore what AI can do for them.

I think seeing it that way has helped to remove lots of the barriers that stop people from using AI in the right way.

Hopefully that’s a differentiator between us and other agencies; we’re firmly seeing it as something that powers a people-led business. We don’t want to be an AI-led business that has people in the background.

And how do you measure the impact of AI? How do you know that it’s actually creating change and delivering results? 

We had two cohorts of people – about 15 people in our Product team and 30 in our Engineering team – complete tasks both with and without AI.

We measured the time taken and the quality of the output. All of this was done anonymously – so we had no idea who had completed each task, and who had used AI.

We found that the time it took to do the task fell by 55%. And crucially, the quality was about the same – or even slightly improved.

Having that data has been invaluable, because it’s helped us build the business case for a larger AI roll-out. And, of course, it’s also allowed us to create this methodology that our clients can use to build a business case for their own AI implementations.

Are there any specific use cases where AI has been particularly transformative for CreateFuture? What makes you most excited? 

Here’s one thing that really wows clients in our AI innovation workshops.

We send all of their in-house experts off to work on a particular business problem.

And at the same time, I run my own “synthetic” workshop and ask AI to dream up its own ideas. I describe the personas of everyone involved in the workshop and get it to predict how they might work together to solve the problem.

Then our real-life experts come back with their answers, and I show them my own AI-generated answers. Usually, the AI comes up with some extra ideas and insights that they hadn’t considered.

And where has it been less helpful? Are there any situations where you tried AI and completely discarded it?

The main thing we’ve seen is that when you rely on it too much, it can get important details wrong – like names and details mentioned in an AI transcription of a meeting. And if you’re using it continuously, those errors multiply over time.

Let’s say I’m transcribing 100 sticky notes from a workshop. I could just take a picture of all the sticky notes on the wall, ask AI to transcribe everything, and then copy-paste the results into a report or a meeting summary.

But AI can make mistakes, even with something as simple as transcription. So I still need to double-check that it’s accurate. End to end, that process still takes me about 45 minutes – but typing up all those sticky notes manually used to take four hours, so 45 minutes is nothing.

It sounds like you’ve put a lot of thought into the cultural and training aspects of AI adoption. Can you expand on that?

It’s basically about creating a collaborative approach that makes sure everyone feels supported and empowered to use AI, rather than overwhelmed or left behind. AI is really exciting, but we have to remember that we’re essentially asking people to pick up a new skill.

That’s why we’ve appointed AI champions in each team – they act as sounding boards and consultants, helping the team approach it in the right way. They help to upskill the rest of the team and signal to people that these tools are here to help them do their best work.

It’s clear that you’ve cultivated a lot of enthusiasm around AI. What advice would you give to companies that are just starting to explore AI but might feel overwhelmed by the process?

Start with small, controlled experiments like we did, and gather data to build your case.

It’s important to understand that AI adoption isn’t just about giving everyone a new tool; it’s about fostering a culture of experimentation and learning.  

By having AI champions, creating a collaborative environment, and focusing on training your people so that AI empowers them instead of replacing them, you can ease the transition and help your team feel confident in using these tools.

If you’re interested in working with people like Angus to develop your own AI adoption strategy, get in touch – and tell us a bit more about your challenge. You can also download our latest guide where we dive deeper into the key insights of our most recent study to unlock the tangible benefits of AI.