My Approach to AI

Love it or hate it, AI is here to stay. There are good uses for it — and some uses that aren't so good.

I've learned how to use the latest AI tools and adapt them safely to my workflow. With the right guardrails, my work is of a higher quality and faster. I'd be delighted to help you safely adopt AI into your business.

It's like a table saw

I think of AI like a table saw. Used properly, you can quickly build quality furniture, or a house for your family, that lasts for generations. Used improperly, you can certainly hurt yourself and others. AI is much the same — an incredibly powerful tool that has to be used safely, responsibly, and sustainably.

How to use it properly

There's a saying: "when the only tool you have is a hammer, everything looks like a nail." I don't treat AI as a universal fix-all — it's one tool in a larger toolbox. Every problem gets assessed individually, and I choose whatever fits: AI, manual coding, or a hybrid of both.

And don't forget the people. AI is genuinely good at the low-level, "non-creative" work that's boring and not the best use of your team's time. Use AI to free your people for the creative, strategic, high-value work that actually needs human insight and empathy.

Good uses for AI

Reading paper forms and doing data entry on them, with human oversight.
Triaging emails and support tickets to identify priority issues and route them to the right person — doing the legwork of research and context-gathering, not writing the reply itself.
Finding specific information buried in large blocks of text — low-level research, always with human oversight to check accuracy.

Can I help you with AI?

Yes. I don't treat it as a buzzword — it's a tool in the toolbox alongside proven traditional techniques. Part of my role at a previous company was developing the safe AI coding policy, covering how AI coding tools were used while making sure customer privacy was maintained, company IP was protected, and every AI output was properly reviewed and validated. At a pathology lab I worked for, I investigated a locally hosted model so the business could get the benefit of a powerful AI system while keeping patient data private. I can help you work through the same questions for your business.

Sustainable use of AI

AI has a real social and environmental cost, so I'm deliberate about how I use it:

Pay-as-you-go over flat subscriptions. Paying per input/output token shows you the true cost of what you're using, and naturally pushes you toward using only what's necessary.
The smallest model that gets the job done — less power, fewer resources, for the same result.
Specific, efficient prompts that reduce wasted compute.
Local models where practical, so you have some say over where the electricity actually comes from.

AI models are improving at an incredible pace, so it's important be able to follow and evaluate new models as they come out, to improve your processes and efficiency over time, while maintaining the same baseline quality.

Privacy and AI

If you use the public AI services - they'll likely use your data to train their models. So don't put any customer data in there!

However, when we get lower level, using AI models via the APIs, the providers have quite clear agreements on the data and what they do with it - so you can carefully use some of these tools with confidence. You must do your research though and ensure you have proper safeguards and processes in place.

As the technology progresses, smaller models, which you can self-host, or run on secure cloud instances, are becoming an increasingly viable option to handle private data.

How I use AI

Programming. I still architect and design every system myself; AI assists by writing the low-level code faster. I review the entire solution — it's my judgement and experience that produces quality software, and I'm responsible for the result, not the model.
Fast research. AI sifts through large amounts of information quickly; I verify and synthesise the findings myself against real references. Microsoft 365 has tools to allow you to safely search through your business documentation and information quickly and answer questions that traditionally would require hours to days of research.
Transcribing paper forms accurately, with a human checking the output — faster and less error-prone than manual typing, without removing the human oversight.
First drafts of documentation and emails. AI gets a solid starting point down quickly; I always review and edit for tone, accuracy, and the human touch.

How I don't use AI

Blind trust. It's a probability machine, generating output based on likelihood, not certainty — and those probabilities aren't always right.
Medical research or safety related queries - it can help point you in the right direction, but ultimately, it lacks the nuance and experience of the correct professionals.
Creative work needing genuine human emotion and empathy. AI can generate art, music, or prose, but it lacks the lived experience and cultural context human creators bring to work that's meant to actually move people.

In summary

AI is here to stay. Like every disruptive technology before it, we — as an industry and as individual businesses — need to work out how to use it safely, responsibly, and sustainably. Should you use it in your business? Yes, if it's the right tool for the job — but understand the risks and benefits properly before you do — and I'm happy to help you work through that.