I built a defence engineering company on precision instruments I understood inside out. I'm not in the habit of using anything I can't take apart and rebuild, and this program is what let me do that with AI.

Before this Quest, my life was three businesses running on Word templates I'd cobbled together at midnight, a filing system that existed mainly in my own head, and a low-grade belief that "AI" meant a chatbot that wrote passable birthday poems. I'd spent years building an engineering company from a workbench to a factory floor, so I wasn't short on rigour. I was short on hours, and I was doing every proposal, every governance document, every client contract the way I'd have done it in 1996: manually, and usually twice. My one act of pure self-indulgence through all of it was a 1975 VW Beetle sitting half-stripped in the workshop, a 2-litre supercharged rebuild that existed purely because I wanted it to, with no business case whatsoever.

The reason I enrolled was straightforward. I'd watched enough of the AI conversation from the sidelines to suspect I was underusing something significant, in the same way you suspect you're underusing a torque wrench when you're still doing the job with a spanner and swearing. I wanted to actually understand the tool, not just prod it occasionally and be mildly impressed. The moment it stopped being theoretical was the night I fed Claude a rough voice note about Engineering Futures, our STEM charity, and got back a fully branded trustee briefing document, correct fonts, correct colour palette, governance structure intact, in less time than it takes me to find my glasses. I remember sitting back and thinking, 'Right. That's not a party trick. That's a colleague.

Where I've noticed the biggest change is in the two places I cared about most. The Change Collective, my NLP and hypnotherapy work, had thirty distinct programmes sitting in my head and scattered across old notes, weight loss, grief, benzo withdrawal, sports performance, thirty conversations I'd had with thirty different rooms of people over years. None of it was written down properly. Now it exists as an actual body of work, documented, structured, ready for the accreditation pathways I'd been putting off because the admin alone felt like a second job. That's years of clinical experience finally on paper instead of trapped in my head like a hostage.

The other is Skyborg, my renewables business. Client proposals, DNO applications, financial modelling, the unglamorous scaffolding that actually determines whether a business survives its own growth, all of it got faster, sharper, and considerably less likely to contain an error I'd only spot at 2am. A twelve-tab financial workbook with over two thousand formulas and not one of them wrong is the kind of thing that makes an engineer unreasonably happy.

My top three takeaways: first, that AI is at its best as a working colleague with context, not a search box you interrogate from scratch each time. Second, that the bottleneck in most of my businesses was never expertise, it was getting the expertise out of my head and onto a page in a form someone else could use. Third, that "I'll document that properly later" is a sentence I can now retire, because later is about twenty minutes away.

Why I'd recommend this program: because it's the difference between owning a tool and actually knowing how to use it. I built a defence engineering company on precision instruments I understood inside out. I'm not in the habit of using anything I can't take apart and rebuild, and this program is what let me do that with AI.

I'm not finished, and the next project is the one that actually makes me grin. For years the engineering, the Beetle, the renewables business and the NLP work sat in separate boxes because I never had the bandwidth to connect them for anyone but myself. The plan now is a YouTube platform that stops pretending those interests are unrelated and puts them in the same place: an engineer rebuilding a supercharged classic on camera, talking through the same precision that goes into a solar installation, occasionally explaining why the human mind runs on the same troubleshooting logic as a misfiring engine. Same voice, same standards, finally one platform instead of four half-built ones. The Change Collective still needs its accreditation pathway completed, Skyborg still has half a business plan to work through section by section, but for the first time in years the documentation isn't the thing slowing me down. It's just work now, and work I can actually get through, with enough left over to point a camera at the Beetle and see what happens.