I get asked about AI a lot, usually in one of two flavours. Either "is this going to run my business for me" or "is it all nonsense." The honest answer sits in the middle, and after using these tools every day to build things for clients around Hull, I have a fairly practical view of where they earn their keep and where they let you down.
Start from what it is good at
At its core, today's AI is very good at producing a plausible first version of something and very good at chewing through text quickly. That is the lens I use for every "can AI do this" question. If the job is to get from a blank page to a rough draft, or to summarise a pile of information, it usually helps. If the job is to be reliably correct with no one checking, it usually does not. Hold that in your head and most of the hype sorts itself out.
Where it genuinely helps a small business
Drafting is the obvious win. Product descriptions, a first pass at a newsletter, replies to common enquiries, a rough outline for a blog post. It gets you to something to react to, which is far easier than starting cold. It is handy for the admin grind too: turning messy notes into a tidy list, drafting a polite chase up email, summarising a long document so you know if it matters. For customer questions, a well set up assistant can handle the same five FAQs you answer twenty times a week, as long as a human picks up anything unusual. And in my own work I use it daily to write and check code faster. Used like this, as a fast assistant that you always review, it genuinely saves hours.
Where it falls over
The trouble starts when people expect it to be right rather than plausible. It will state something false with total confidence, so anything that has to be accurate, prices, opening hours, legal wording, a customer's actual order, needs a human to check before it goes out. It does not have your judgement or your history with a customer, so it cannot make the call you would make. And it does not sound like you unless you push it hard, which matters, because the thing that makes a small business worth choosing is usually the person behind it. Hand your whole voice over to a machine and you sand off the bit customers came for.
The sweet spot: assistant, not replacement
The businesses getting value out of AI are not the ones trying to replace people with it. They are the ones using it to take the boring first 80 percent off a task so a person can do the last 20 that actually needs a human. Draft, then edit. Suggest, then decide. Keep a person in the loop on anything a customer sees or anything that has to be correct. That is not a fashionable take but it is the one that holds up in real work.
What I would try first
If you run a small business and want to test this without betting the farm, pick one annoying, low risk job and try it there. Drafting replies to routine enquiries. Turning your scribbled notes into a first draft of a page. Writing product descriptions you then edit into your own words. Keep it to things where a wrong answer is a quick fix, not a disaster, and see whether it actually saves you time. Most people find one or two jobs where it fits and a lot where it does not, and that is a perfectly good result.
A word on the bigger stuff
There is a heavier conversation about AI policy, data, and governance once a business starts using these tools properly, especially around what you feed into them and who is accountable for the output. That is worth doing, but it is its own topic and a bit beyond a freelance web developer's remit. My other site covers the strategy and governance side in more depth over at headless-digital.com, so I will point you there rather than pretend it fits in one blog post.
If you would rather just talk through where AI might sensibly fit in your setup, without the sales pitch, that is exactly what my AI consultancy help is for, and you are welcome to grab a free thirty minute surgery on the contact page. I will tell you honestly where it will help and where it will waste your money.
Free 30-min surgery
Working on something similar?