Ai can now be applied to almost any business task. But does using it actually makes that particular task any better?
Another in our series inspired by recent conference presentations. There’s a link to the original video at the end.
There’s an old saying that when all you have is a hammer, everything starts to look like a nail. Ai may be giving businesses a very large and impressive hammer.
Once we’ve discovered that it can write an email, summarise a document, analyse some data or produce a first draft in seconds, it’s tempting to start wondering where else we can use it. Before long, the question has changed from “Would Ai improve this?” to “How can we use Ai for this?”
There’s nothing wrong with experimenting. In fact, experimenting is probably the best way to discover where Ai genuinely helps. But it works much better if we start by deciding what we want from it? Doing the job faster? More cheaply? More accurately? Or simply remove a tedious task from somebody’s day?
Without that test, almost any Ai experiment can be declared a success, and marketing offers plenty of examples. Ai might let us create ten social posts in the time it used to take to write one. That’s impressive if those ten posts achieve more than the original one. It’s considerably less impressive if we’ve merely become ten times faster at publishing material nobody particularly wants to read.
The useful habit, then, may be to treat Ai as something to test rather than something to introduce by default. Try it on a defined task. Decide beforehand what ‘better’ would mean. Compare the result with the existing approach. Keep it where it clearly helps, adjust it where the result is mixed, and abandon it where it doesn’t.
Importantly, an unsuccessful experiment isn’t wasted effort. Discovering that Ai isn’t particularly good at one part of our business is useful information. It stops us investing more time in an idea simply because everyone else appears to be doing it.
The smartest Ai users may eventually turn out to be the people who’ve worked out where not to use it.
- Inspired by the conference presentation “Lost & Found in AI Wonderland: An Honest Journey Through Hype, Headaches, and Real Wins” by Nirav Mehta at WordCamp Asia 2026 – watch the full video here.
