As the founder of General Purpose, one of the UK's leading AI training companies, I've had the chance to work with hundreds of firms as they've started their transition to becoming AI-native. The interesting question is no longer which AI tool to roll out. It's whether your firm can become AI-native fast enough to matter.

By AI-native I mean a firm where AI is built into how the place actually works, not bolted on the side. The gap between firms doing the work and firms still talking about it is widening fast.

Here's what we're seeing across our clients.

Chefs in Schools is a charity transforming food in UK state schools. Small team. No tech infrastructure. Shoestring budgets. Every minute of the day spoken for. They've made AI a routine part of how their team works. If they can do it, the "we're not really a tech business" excuse is finished.

Cowshed are another good example of what AI-native looks like in a non-tech firm.

Their head office team spent a day with us working out where AI could fit across the business. Then they went back to their desks and started building. One wired ChatGPT into Outlook to get a live view of outstanding orders. Another built a skill that lets them quickly check which products include which ingredients. A third replaced a multi-spreadsheet hunt with an agent that fills out retailer info forms automatically.

They are, quite literally, building AI capabilities their competitors don't have and can't replicate. They're using their people's knowledge and willingness to experiment as a moat to build a new form of competitive advantage.

And it's not a one-off.

We recently worked with Frontier Economics, the European economics consultancy, to shape their AI strategy and upskill their team across London, Paris, Berlin, Madrid and Dublin. Again we saw a cambrian explosion of people building the tools and capabilities they need to get more done.

An ESG questionnaire drafter that does in minutes what used to take half a week. A monitor that tracks new UK competition cases and pings the relevant team. A weekly digest of German energy policy changes that lands in front of the partners advising on them.

None of those came from a strategy roadmap. None came from a vendor pitch. They came from individual economists looking at their own week and thinking "this is the bit I'd love to hand off". Multiply that across hundreds of consultants trading what's working and egging each other on and you've got a firm that's getting sharper every week.

But none of this is possible without laying the foundations first.

An engineering firm we trained earlier this year saw how even an industry as traditional as bridge building could be accelerated with something as simple as giving people Copilot and getting them to actually use it. It hasn't got the glamour of end-to-end automation but in terms of business impact, broad adoption quickly adds up.

And some of the most ambitious use cases depend on work that doesn't involve AI at all.

Access Holdings, a US private equity firm we work with, have spent the last few years figuring out how to pull transaction-level data from 250+ portfolio company locations into one central platform. This was a complex standardisation and negotiation process that had far more to do with people and change management than the tech that underpinned it.

However, once it was done it suddenly became a piece of cake to add an agent that highlights unexpected performance or spots opportunities for expansion. Now they're saying that their ability to interpret that data on the fly is a key part of their competitive advantage.

And closer to home, at General Purpose we've spent the last year wiring our own systems together, allowing our team to use AI to query and update our CRM, accounts, calendars and our learner platform.

I can ask "who in our pipeline needs a follow-up, based on when we last contacted them and what we last said?" and get a prioritised list with all the info I need to pick up the phone. I can ask for next month's profit forecast and then to increase our advertising spend based on the predicted ROI. I can ask for a new class to be set up for a client and an instructor provisioned on the platform, all while waiting for the bus.

The firms we see succeeding are those who are investing in making AI their edge. They're investing in their people - making sure they have the mindset and skills to take advantage of AI. They're investing in cleaning and connecting their data. And they're investing in getting the governance right - making the hard choices about what they're going to de-prioritise in order to create space for change. 

And if you'd like to see what that looks like in practice, come and take part in the Builders Track of the AI Skills Accelerator Training on 9 June at Studio Spaces.