I was talking to a close family member one day. Just chatting, nothing important. And somehow we got onto AI.
"I don't use AI. I think it's cheating."
They weren't being awkward. They weren't being a technophobe. This is someone bright and capable who picks up whatever they need to. But somewhere along the line they had decided that using AI to help them think or write or work would be a kind of fraud. As if leaning on it would mean the work wasn't really theirs.
It was such a small moment. But it stayed with me, because I knew even as they were saying it that millions of people feel exactly the same way. They watch AI from the other side of a line they've drawn for themselves, and they call it cheating, or lazy, or not-really-for-people-like-me. And they quietly opt out of something that could change what's possible for them at work and in life.
The moment
So I started to explain it. Properly. Not as a tech sell — as a family member trying to help someone I love see it differently.
And somewhere in the middle of explaining, I heard myself say something I hadn't said before:
"Don't think of it as artificial intelligence. Turn the initials around. Think of it as your intelligent assistant. You're the intelligence — it's just helping."
AI → IA.
Same technology. Completely different relationship with it.
I didn't plan to say it. It came out because I was reaching for language that would land for them — for one specific person sitting in front of me, with one specific anxiety about technology. The acronym flip wasn't a marketing exercise. It was a reach.
But the moment I said it, something shifted. They got it. The thing they'd been resisting wasn't the technology — it was the framing of the technology. The word "intelligence" sitting in front of "artificial" was doing all the heavy lifting in their resistance. The minute they could see themselves as the intelligence and AI as the assistant, the cheating concern dissolved. Of course it's not cheating to ask an assistant for help. That's literally what assistants are for.
I'd already been living it
That conversation didn't feel like a business idea at the time. But it stayed with me, partly because I'd already been quietly living what I'd just said, without realising — for a year before, and in a different setting again at the moment we were talking.
Let me start with the year before.
I'd been brought into a BAU Scrum team enhancing a complex legacy pricing model. The kind of system you find at the heart of a lot of established businesses — Excel and VBA, decades of accumulated logic, one or two people who really understand it, and the business runs on what it produces.
As part of my consultancy contract, I had a Copilot Pro licence. And over the months of that engagement, I'd ended up restructuring almost everything about how I did the role.
Release notes used to take me hours every sprint. With Copilot, I drafted them from the delivered stories in a fraction of the time. The judgement about what mattered was still mine. The slog of putting it into language wasn't.
Backlog refinement was the same. Going into a planning session with a clean, refined backlog used to mean an evening of work the day before. Copilot helped me prepare it from Jira straight into something the team could actually engage with. I still did the thinking. It did the typing.
And the team had some quirky internal processes — the kind of fiddly admin every team accumulates over time, that nobody loves but nobody can quite get rid of. I automated several of them in a few afternoons. Years of friction, gone.
I was directing the tool. I was the intelligence. It was the assistant. I was reviewing, refining, redirecting. I was holding accountability for the output. I was using it to do things I genuinely couldn't have done at the same time before.
The hours that freed up didn't go into a coffee break. They created the capacity for me to take on a second Scrum Master role on a complex cloud migration, running alongside the pricing-model work. The work I could do had expanded — not because I worked longer hours, but because the relationship with the tool had changed what was possible in the same hours.
The meetings I didn't attend
There's one detail from the cloud migration that captures the whole thing for me.
I was the only person across that migration with a Copilot licence. The other teams didn't have access. They were running their meetings the way everyone runs meetings — someone trying to take notes, someone else trying to participate, actions getting lost in chat logs.
So I started hosting meetings for other teams that I didn't even attend. They'd invite me as the host so Copilot Recap would run. After the call, the notes and actions would land in the chat, and the team could get on with the work. I'd never opened my microphone. I'd never even seen the agenda.
Looking back, that's the framework in miniature. Same technology, used by a person who saw their relationship with it differently. Service extended to people who didn't yet have the tool, because someone with the tool was thinking about how it could help them.
A longer view
That engagement wasn't the first time I'd worked like this. It was just the first time I noticed.
Before consultancy, I spent 14 years at Nelsons in Wimbledon — a UK natural health care products manufacturer working under a pharmaceutical manufacturing licence. For most of it, I built and ran an integrated systems landscape that I'd architected from the ground up. Sage Line 500 (originally Tetra CS/3) as the ERP backbone. A warehouse management system integrated alongside it. Preactor for demand management; TXT for demand forecasting. Behind all of that, the data warehouses I'd built, reported on through Business Objects. The year before we moved to SAP, we implemented BPC for financial reporting, forecasting and budgeting. And in the closing years of my time there, as the business outgrew the system that had got it there and the regulatory environment we were heading into demanded more, we migrated the whole estate to SAP itself. I led that migration. At go-live, we delivered at less than 0.1% error per million rows — against IBM's forecast of 10%.
I mention all of this for two reasons.
First, because regulated manufacturing teaches you something specific about technology that I think most AI conversations are missing. Tools matter, but the people running them matter more. You can have the most expensively-licensed software in the world; if the people using it don't trust it, understand it, and know when to override it, you have a compliance risk dressed up as a capability. The same is becoming true of AI in any organisation that's serious about getting value from it.
Second — and this matters more for what I'm building now — data has always been my thing. The IA Framework principle that the quality of the output reflects the quality of the brief is something I learned years before AI existed in any usable form. When you spend your career building data warehouses, you internalise the truth that intelligent systems are only ever as good as the structured input you give them. AI is exactly the same. The reason most organisations get poor output from their AI tools isn't because the tools are weak. It's because the briefs are.
Then I was made redundant
The contract ended in the way these things often do. I was made redundant. After thirty-five years in technology — IT manager, software consultant, Scrum Master, ERP and CRM and BI delivery — I was suddenly out, and I had to decide what to do with everything I'd learned.
The honest answer is: for a while, I followed a different path entirely.
The obvious move
What my CV pointed at was clear. Set up a traditional project management and Agile delivery consultancy. That's the work I'd done for two decades after the pharma role. That's where my credibility, my network, and my proven track record sat.
So I researched it. Properly. I spent weeks with Perplexity working through models, niches, propositions, pricing, target sectors. The outcomes were genuinely interesting. There were several routes I could see a path through.
What I didn't have was the energy to take any of them forward.
It wasn't imposter syndrome. The path was clear enough; the appetite to walk it wasn't there. I'd done a version of that work for thirty years. The thought of running it back, with me as the brand instead of the firm, didn't move me.
I sat with that for a while. The obvious move that didn't move me. It was an uncomfortable place to be.
What I was doing instead
While I was sitting with it, I was also doing something else with my time.
I was building.
I set up a homelab — the kind of personal infrastructure project that engineers do for fun, and that I'd never had the time for during three decades of client work. I built a WordPress theme. Then a WordPress plugin. Each one a thing I would have considered, before all this, properly out of reach for someone with my background. I'm a delivery, BA and data person, not a developer.
But I wasn't building alone. I was building with AI as my assistant, applying — without yet having a name for it — exactly the same principles I'd been using on the pricing model and the cloud migration, and at Nelsons years before. Direct the tool. Review the output. Refine. Hold accountability. Expand what's possible.
By the time I'd shipped the plugin, something had changed. I had proof. I'd watched myself become someone capable of shipping things I genuinely couldn't have shipped alone. Not because the tool was magic, but because the relationship with the tool was right.
That proof was what the research couldn't give me. The research could outline a consultancy. It couldn't generate conviction. The building did.
And around the same time, the conversation
It was around this period — concurrent with the building, the not-acting on the obvious move — that I had the conversation I opened this post with. The "I don't use AI. I think it's cheating" conversation with a close family member, where I reached for the words that would later become the spine of everything.
In retrospect, those two things were happening in parallel for a reason. The building was showing me, with my own hands, what was possible when the relationship with the tool was right. The conversation was giving me the language to describe what I'd been doing. The lived experience and the words finally met.
What happened next
After that conversation, something started happening that I noticed but didn't yet have a name for.
Whenever AI came up — with friends, with family, in passing — I'd find myself sharing the same reframe. You are the intelligence. It is the assistant. And every single time, it landed. People got it. They asked follow-ups. They started thinking out loud about what it would mean for them, for their work, for their team, for the relative who'd dismissed AI as cheating.
I wasn't market-testing a proposition. I was just sharing what I'd come to. But the response was unmistakable.
Somewhere between the building, the conversation, and those follow-up conversations, the answer arrived. The consultancy I should be starting wasn't the obvious one. The thing I'd been doing all along — the thing that came naturally, the thing people lit up over — was something else.
That's where Mv32 came from.
A note on where the servant heart comes from
There's a strand running through all of this that I should be honest about.
Alongside my consultancy work, I've supported a small charity for years — administering their Microsoft 365 estate since before Covid, and contributing across governance and operations in different ways. Quiet work, mostly. Not paid, not public.
But it's the same instinct that runs through everything else I've described. Helping people use technology in ways that actually make their work and their lives better. Not pushing the tools. Helping the people. That instinct didn't start with the IA Framework. The IA Framework is the language I've finally found for an instinct I've always had.
From a sentence to a framework
That one line — you are the intelligence, it is the assistant — became the spine of Mv32.
It became the IA Framework. Five principles to live by:
- 1.You Lead, It Follows. Your expertise directs; the assistant executes.
- 2.Brief It Like a Professional. The quality of the output reflects the quality of the brief.
- 3.Work in Loops, Not Lines. The first output is a starting point, not an endpoint.
- 4.Trust But Verify. AI is confident even when it's wrong; you remain accountable.
- 5.Expand What's Possible. The real value isn't doing the same things faster — it's doing things you couldn't do before.
It became a coaching arc — five phases that move people from Awareness through Direction, Collaboration, Integration, and finally Expansion. Most teams I work with arrive somewhere around Phase 1 or 2. My job is to move them forward deliberately, not throw them into the deep end.
And it became a question I come back to in every session, with every client:
"You are the intelligence — so what do you want your assistant to do?"
The third story
Here's the bigger thing I want to say, and the reason I'm now building this in public.
There are two stories about AI in circulation at the moment.
The first is the press story. The robots are taking over. AI is coming for your job. Be afraid. This story sells papers and gets clicks, but it doesn't help anyone do better work or build better organisations. It just makes good people anxious.
The second is the licence-and-hope story. The UK Government's own AI Adoption Research, published earlier this year, found 71% of organisations haven't identified a clear use for AI — and 60% cite skills and expertise, not budget, as the main blocker. That picture matches almost everything I see in the SMEs and charities I talk to. Copilot or ChatGPT seats bought, used occasionally for drafting emails, and six months in nobody can point to anything that's fundamentally changed in how the work happens. The licences are paid for monthly. The work, fundamentally, looks like the work did before.
I think there's a third story. A more honest, more useful one.
AI, properly framed, isn't the job-taker. It's the time-giver. It frees teams to do the more interesting, more rewarding, more impactful work they've never had time for. It lets one person host the meetings, run the migration, ship the plugin. It turns the friction in everyone's week into capacity for the work that actually matters.
But this only happens when people relate to the tool the right way. When they're the intelligence, and they treat it as the assistant. When they direct it, brief it, refine its output, verify what it produces, and expand what they can do.
That's the change of view I'm trying to put into the world. Not "AI will help your business." Not "AI will replace your team." Something more useful and more true: AI will free your team to do the work they've always been capable of, if you frame your relationship with it right.
Why I'm doing this
I'm not building Mv32 because the world needs more AI training. There's plenty of that already, and most of it starts in the wrong place — with the tool.
I'm building it because somewhere right now, someone like my family member is quietly deciding AI isn't for them. They're calling it cheating, or hype, or not-really-for-people-like-me. Somewhere else, an SME owner is paying for Copilot seats and quietly wondering whether they're getting any return on them. Somewhere else again, a charity is being told they should be using AI and doesn't know where to start.
I'd like to help with all three.
That's the work.
A question for you
I'd love to hear from anyone reading this — particularly anyone who has had a conversation like the one I had with my family member, on either side of it.
Who in your life has said something like "I don't use AI, I think it's cheating"? And what did you wish you'd had the words to tell them?
If you've found those words, share them. If you haven't, maybe this post is them. Either way, drop a comment — I read every one, and I'd love the conversation.