Human error drives me nuts. It should not happen. But it happens all the time.

When it upsets customers, it becomes a big problem.

Now I know I always tell you that my business runs like a well-oiled lean machine. And it does most of the time.

But there is no business on the planet that is 100% well-oiled all the time. Problems will arise, customers get grumpy and sometimes you will be wondering and asking yourself ‘how the heck did we allow that to happen.’

I have a team of 11 people. What are the chances that as I write this sentence one of them is not undertaking a monumental screw up? No matter how good your systems, human error will ensure they fail.

And so it happened to me on a Sunday. I get a call from a new client who has opened an email from my team (which is automated in the background) stating they needed to pay $94,000 within the next few days which was totally unexpected.

And this type of news is not something you get grumpy over. It’s the sort of news that can send a client into panic mode.

Now I pride our firm as one that gives clients peace of mind and a feeling they are in control. You could say we missed the mark a bit in this case. To me it was like shooting an arrow from 50 meters with target of a pin head. Now, let me be clear. The BAS work we had done was correct. The amount due was payable. It was just that the issue came about because of something the previous accountant had done. The issue was that despite having procedures that state we should call clients and explain the situation and giving options on how to deal with it, we just sent out an email which at the best of times is impersonal.

But once explained and given some options it made sense to our client. But that did not stop the panic in the first place.

But this got me thinking. Should I just sack my team and start using AI? Or if I cannot trust them, should I do all the work because no one is better at this than I.

Well, if I go for the later, it will mean I work all hours of the day and must burn the midnight oil too. And if you know me you will know that is not an option. And it should not be an option for you as a business owner either. To do it this way is to guarantee long working hours making your business totally dependent on you. And trust me, you will make errors just like your team.

AI then? After all they say it is about to take over the world. There are a couple of issues with AI as we all know. One of which it tends to make things up. Besides despite the above I love my team and what happened was out of character but then again if I could eliminate human error should I not take advantage? And save some money in the process?

So, I went into full blown research mode. Much science was involved too. To be honest there was no science and the research involved Google which we all know is not exactly accurate either.

Anyway, it appears AI can be used for BAS analysis. It may not be able to do a BAS, but it can interpret one. It can find things that are unusual and given the right parameters it can bring certain matters to the attention of the user. And going forward we might just use it.

But this is what the whole episode taught me. It does not matter if you use a human or use AI. They will both get it wrong sometimes. And you need to have the right mix. The right balance between human interaction and machine interaction.

The ability for AI errors may be reduced but we humans like to deal with humans. Even if you are a millennial when you do not understand something or need clarification you want to talk to a human voice not an AI bot. AI will, as we introduce it, take over more mundane work. It will be able to do things in seconds that take a human an hour. But it will still make mistakes. Just like a human.

And as an employer, you must decide how much you are prepared to automate and how much of a human touch is required.

For now, I think we will stick with the human touch with as much automation in the background as clients can handle. And I have told my team that screw ups are not acceptable but at the same time told them they will screw up. It’s human nature to screw up but our job is to limited screw ups as much as possible. At least I can tell them that.

I can try telling (or shouting) at AI/ your PC in the hope it will change (I have tried this many times) but it never works.

And if like me you like to know about your team’s personal lives, what their children are doing and what they did over the weekend then good luck with asking AI that. I did and was simply told ‘I do not have personal experiences or consciousness as I am a machine.’ And that is probably the real reason we might wait a bit before going full blown AI.