By Catherine Pretorius
AI is extremely good at getting things done quickly.
It takes only a few seconds to draft an email, meeting notes can be summarised almost immediately, research can be put together more quickly than most of us can brew a decent cup of coffee, and repetitive administrative tasks can be delegated to systems that never mind about Mondays.
That’s helpful, particularly in the case of small businesses in which a single person may have to look after clients, draw up invoices, carry out marketing, provide customer service, plan things, and at the same time encounter some technological problem that shows up at the most inconvenient moment; AI can then seem like an extra pair of hands.
Here’s a question that’s worth asking.
Does the fact that AI helps us carry out all activities more quickly automatically mean that we are carrying them out better?
Faster Is Valuable, but It Isn’t Everything
Clients value quick replies; owners want information without waiting 3 days for someone to compile it; and no one misses the days when the same task had to be repeated manually 50 times because “that’s just how the system works”.
AI can reduce friction by organising information, preparing the initial documents, detecting patterns in data, responding to standard questions, and performing tasks that don’t require human attention.
That frees people, or at least that is the idea.
The danger is that every bit of time saved opens the door to even more work, more messages, more tasks, and greater pressure to respond at once, because it is now technically possible to do so; efficiency can therefore become a trap.
AI Is Brilliant at the First Draft
A major advantage of AI is its ability to get things off the ground.
Facing a blank page can be intimidating, especially if you’ve already spent this morning dealing with a billing problem, replying to emails, and trying to recall whether the thing you said you’d do “later” was actually due yesterday. AI helps to reduce that resistance.
It is useful for creating a structure, making sense of a confusing idea, producing a first draft, or giving you a solid starting point. This can be very helpful, but it should be remembered that it is still just a first draft.
It doesn’t always point out the detail most important to your client, the sentence that doesn’t sound like you, or the situation in which the technically correct answer isn’t in fact the most sensible one.
That part still has to be handled by a human.
Research Is Faster, but Judgement Still Matters
AI can also speed up research.
It can summarise large quantities of information, compare different ideas, and detect themes, and thus will enable you to get up to speed on a particular subject without having to spend the entire afternoon browsing the internet aimlessly.
It is indeed a real advantage, but the fact that information can be accessed quickly doesn’t mean we don’t have to check whether it is accurate, up-to-date, relevant, or appropriate. Artificial intelligence may seem confident even when it is wrong.
Of course people can. We’ve had centuries of experience.
What this shows is that even if a company employs information produced by AI, its obligation to verify that information does not disappear; indeed, it becomes all the more important since the output can be refined in a manner that causes people to cease questioning it.
It’s useful to have convenience, but it’s not useful to have blind trust.
Customer Service Is Where the Difference Becomes Obvious
AI can do a great job answering simple customer questions.
It can handle common questions, lead people to the correct information, and cut down on the amount of time taken to respond each time to the same simple request.
It does make sense for it to be about routine support, but customers aren’t always routine.
There are situations in which the person contacting you is frustrated, confused, worried, or encountering a problem that cannot easily be dealt with by using a standard response; in these cases, it is more important to understand than to be quick.
An answer given quickly that doesn’t get to the point isn’t always useful; in fact, it’s more likely to gain people’s trust if it comes from someone who really understands the problem and who takes the time to give it.
It’s the way people are treated that they remember, particularly when things go wrong.
That doesn’t mean that AI should have no role at all; instead, there should be a point at which the system recognises that a person needs to take over.
Better Business Still Requires Context
Where AI operates, there is usually context in business.
A customer who usually pays on time might receive a different reaction from one who consistently ignores invoices. A long-time client who is facing a real emergency might require more flexibility than a normal policy provides. Although a solution may be technically valid, it could still cause an unnecessary problem somewhere else.
The decisions aren’t always difficult since the relevant information is missing but rather because the information has to be interpreted; that is what judgement consists of.
Even though AI can help in making a decision by providing information, showing patterns, and highlighting risks, the final decision is still greatly influenced by experience, instinct, empathy, and an understanding of the people involved.
There isn’t a simple algorithm which can be applied to all the grey areas.
Relationships Don’t Scale Quite So Easily
The promise of automation lies in its ability to achieve scale.
At the same time, you will be able to communicate with more people, handle a greater number of enquiries, produce more content, and cope with a higher level of activity, all without increasing your workload.
That’s attractive.
That’s not the way relationships work.
People can detect the fact that all the emails have the same tone; it’s noticeable when a reply refers to the wrong point, and it’s clear that one is talking to a system that has no awareness that the problem has already been explained three times.
A lack of personality in a business can make it seem rather remote.
That is significant because a large number of small businesses operate in fields which larger organisations find difficult to automate—fields involving familiarity, flexibility, trust, responsiveness, and the feeling that someone actually knows who you are.
The greater the access people have to the same AI tools, the more valuable human qualities will be.
The Risk of Becoming Too Polished
AI can make businesses look more professional. That’s generally a good thing.
Poorly written emails can be improved, awkward wording can be polished, and as a result communication can become clearer and easier to understand. But there comes a stage where all things begin to sound identical. It’s well organised, properly polite, and entirely forgettable.
A business voice shouldn’t be messy simply because it wishes to appear authentic, but it also shouldn’t sound as though every sentence has been produced by the same corporate polishing machine.
People respond to personality.
A little humour, warmth, directness, or even a certain awkwardness can make communication appear to be genuine.
The term ‘professional’ doesn’t have to imply that someone is robotic.
In a rather ironic way, AI is making that distinction more significant.
What should we do with the time that AI frees up?
This could be the more interesting question.
What happens to the five hours that AI saves a business each week?
Do they then immediately disappear into another pile of tasks?
Could some of that time be put to use in having better conversations, planning more effectively, thinking more deeply, in improving service, in staff development, or simply in reducing the persistent feeling that everything has to be urgent?
It’s always been claimed that technology can save time. Perhaps the goal shouldn’t simply be to do more but to do the important things better.
AI Should Remove Friction, Not Responsibility
There’s a tendency to believe that once something is automated, it’s no longer really our problem.
It is.
When an AI system sends the wrong message, provides bad advice, misunderstands a request, or produces inaccurate information, the customer won’t blame the algorithm; instead, they will blame the company using it.
That’s reasonable.
Although AI can help with decision-making, communication, research, support, and administration, responsibility still rests with humans.
That is, in fact, the most important distinction of all: make use of the technology but continue to take responsibility.
AI will remain part of business as it becomes more capable, more widely used, and probably a great deal less noticeable as it is integrated into the tools we already use every day.
It’s not something that businesses have reason to fear.
If used properly, AI can eliminate dull tasks, help people find information more quickly, improve communication, and give small teams capabilities that have heretofore mostly belonged to large organisations.
But speed is only one measure of progress.
To provide better service, it’s necessary to understand. For better decisions, judgement is needed; for better relationships, trust is required; and when it comes to better businesses, people who care about the consequences of the choices made are still needed.
So go ahead and let AI speed up business.
Just make sure the time, energy, and attention it gives back are being used to make the business better too.