Automation with AI Chatbots – a support tool or an unnecessary fad?

Good evening.
My name is Maja, I am part of the team — and I take care of clients from Slovenia.
I have analyzed your website…

This is how an email I received last week began.

And if this were a real person, I would say – okay, she made a typo. But “Maja” is not a person. It is an AI solution that the company offers as a “virtual human agent.”
And it is precisely this “I have analyzed” that is an excellent symbol of the current state of artificial intelligence: confidence without understanding.

You won’t believe it, but Maja still manages to sell her solution. Her customers think:

“Let the errors remain in the text. So that it is clear that we are using artificial intelligence.”

This is a mindset where the goal of digitalization is not to improve a process, but to leave traces of digitalization. Not for AI to help humans – but for humans to prove they have AI.

Such a director does not understand that artificial intelligence without human intelligence becomes merely a generator of random impressions. And that technology that makes mistakes is not a sign of progress, but a sign that the company has no one to notice them.

Artificial Intelligence or Artificial Confidence

With the development of AI, many entrepreneurs fear they are falling behind, missing opportunities. That if they don’t “keep up with the times,” they will soon become irrelevant.

One of these “opportunities” is also a website chatbot. “An AI chatbot that communicates with customers on its own,” we hear. It sounds impressive. But the question is: who actually needs it?

First condition: volume

These “things” learn from mistakes and examples. If there aren’t enough conversations, there’s no learning.
The average Slovenian company, with a handful of visitors and a few inquiries per month, will not have enough data for the system to even “improve.”

For them, an AI chatbot will be more like a mannequin in a shop window – something that looks modern but does nothing useful.

Second condition: price that grows with “intelligence”

At first glance, everything seems cheap.
A website, one plugin, a monthly subscription of 20 euros.
But when we want the chatbot not to talk nonsense, additional costs arise:

• system configuration,
• communication tone adjustment,
• entry of frequently asked questions,
• payment for the number of requests (API calls).

With a larger number of conversations, the bill also grows.
Then we are back to the same question as with advertising: can it be cheaper?

Learn from the big ones – but don’t imitate

The largest companies have implemented AI chatbots primarily in after-sales support, not in sales.
Why? Because automation does no harm there.

If the chatbot answers:

“Have you already checked if your internet connection is working?”

this is not annoying, but practical.

However, if it starts explaining something unrelated to the customer’s question in the middle of a sales conversation, it is a lost opportunity.

Good practice is for the bot to ask a few questions at the beginning, such as:

“What are you interested in?”
“Are you already using our product?”
“Have you tried restarting the device?”

and then – hands over the conversation to a human. This person enters the conversation somewhat prepared, with basic information already “on the table.” Some who forgot to “turn on” the internet have already solved their problems. Thus, mostly only those who truly need an expert reach them.

In this case, the customer gained something by getting faster interaction and problem resolution, and the provider gained something by already filtering urgent from non-urgent cases.

This is how technology helps, rather than replaces.

When a company implements AI because “everyone else is doing it”

One of the biggest problems is not in the technology, but in the mindset.

Companies implement AI solutions not because they need them, but because they have simply appeared.

They feel they will miss the train of the future if they don’t have “something artificially intelligent.” But a train going in the wrong direction is no better than standing still.

The truth is that AI solutions are effectively used only by those who already had well-organized customer communication.

Those who have ignored chats and inquiries for years will not miraculously start processing them just because “artificial intelligence” is now involved in the conversation.

Technology is not a solution if there is no need

An AI chatbot has its purpose. But only where there is a need for conversation, not just a desire for modernity.

For someone with thousands of interactions monthly, it is a tool that can:

• shorten response times,
• retain dissatisfied customers,
• filter basic questions.

For a smaller entrepreneur, however, it is often just another plugin that will quietly disappear from the website after a few months.

Smart people, not smart systems

The greatest danger of artificial intelligence is not that it will be smarter than us, but that it will give confidence to those who do not understand how to use it at all.

AI chatbots are not a solution for poor communication – they are merely a magnifying glass that reveals it. And until Maja learns to write a message without spelling errors, there’s really no point in trying to analyze my website.