Are We Really Lucky to Live in the Age of AI?

The question is rarely asked out loud, and almost always with a note of worry: are we really lucky to live in the age of artificial intelligence? It deserves better than a reflex answer, in either direction. Here is ours, and why it leans optimistic.

Fear has come with every tool that mattered

The printing press was going to dull readers by sparing them the effort of copying. Electricity was going to unbalance the body. The pocket calculator was going to kill mental arithmetic. The internet was going to replace shops, then newspapers, then school. The same worry each time, stated just as seriously, by intelligent people.

It would be dishonest to conclude those fears were ridiculous. Some transitions genuinely hurt, and hurt specific people rather than an abstraction. But what disappeared were tasks, not whole trades: the typesetter became a designer, the accountant stopped adding up and started advising. The tool pushed the effort upwards.

What AI actually changes for a small business

This is where the era is frankly favourable, and that is new. For thirty years, computing mostly served organisations that could afford a team. Today's tools do the opposite: they put within reach of one person what used to require several trades.

Translating a website into two languages without a translation agency. Drafting a first version of two hundred product descriptions. Answering a customer who writes at eleven at night. Finding, in six months of messages, a commitment made in writing. These are tasks that used to cost time or money, and now cost almost neither.

For a craftsman, a shopkeeper or a team of a few people, the gap with large companies has never been this narrow. That is good news, and it is rarely told as such.

What it does not do, and this needs saying

A model does not decide. It proposes, with identical confidence whether it is right or wrong, and that is exactly what makes it misleading. It invents references that do not exist, gets a figure wrong without flagging it, and words its mistake as well as its correct answer.

It carries no responsibility either. If a generated text commits your business, your signature is at the bottom of the page, not its. Judging, deciding and checking remain our work, ours as humans, and they will for as long as a human answers for what is published.

That is why being optimistic does not mean being naive. The tool is remarkable as long as we are the ones using it. It becomes risky the day we simply apply what it says.

And for the generations that follow?

That is the real question behind the first one. Our children will grow up with a tool that answers everything, instantly, and never says it does not know.

The risk is not that they will know less than us. It is that they will stop checking. Knowing where a piece of information came from, when, and whether another source confirms it: that skill has never mattered more, and it is not learnt on its own.

But look at the other side of the scale. A curious teenager now has, for free, an infinitely patient interlocutor who will explain a concept ten different ways until it lands. A struggling pupil can ask the same question three times without anyone sighing. A child far from any large city gets the same explanations as one schooled in the capital.

That equality of access to knowledge has never existed in human history. Not once. If you are looking for a reason to be optimistic, it is there, and it is hard to overstate.

So, are we lucky?

Yes. On one condition, and it is not a technical one: that we, as humans, keep the decision. The model proposes, it signs nothing. It is our name at the bottom of the page, never its own.

AI amplifies what we bring to it: care as readily as laziness, curiosity as readily as naivety. It does not choose, we do.

Our children will inherit a remarkable tool, and the duty to learn to check it. That is demanding. It is also, all things considered, a far better problem than what history left to most generations before us.

How we use it

At Devoptus, AI enters a project where it delivers a measurable service, not to tick a box. A conversational assistant connected to your site content answers your visitors' recurring questions at any hour, and tells you what they look for without finding. It is a costed deliverable with a scope written into the quote, not a promise.

What we do not promise, on the other hand: making your site appear in the answers of one assistant or another. That ranking belongs to whoever publishes the model, and nobody else can commit to it. Being wary of anyone selling you that is part of reasonable optimism.