A year inside artificial intelligence, told by those who use it every day – and continue to sign what they produce.
We're asked about it at dinner, among friends and family, with that mix of curiosity and apprehension that accompanies everything new. “Aren't you afraid of the new tools coming out every month?” “My friend got his logo made with ChatGPT, and it turned out well.” And, beneath each of these phrases, the real question, the one few are willing to fully voice: will artificial intelligence replace us?
It's a legitimate question, and it would be dishonest to dismiss it with a joke. But, after hearing it asked so many times, we've come to a conviction: it's the wrong question. Not because it doesn't have an answer, but because it puts the wrong thing at the center — the machine. And when the machine is at the center, every answer is already flawed from the start. Those who fear it imagine it as a replacement; those who worship it imagine it as a miracle. They are two sides of the same mistake: both have removed humans from the center of the conversation.
It's worth remembering that this isn't the first time. When photography arrived, many painters announced the end of painting; painting, instead, stopped having to resemble reality and discovered everything else. The digital tablet didn't manage to kill painters either. The calculator hasn't eliminated mathematicians. The tools that seem to threaten a trade almost always just shift its center of gravity: they remove the mechanical labor and return the part that matters. Our bet at Demetra is that the same will happen with artificial intelligence—with one condition. Yes, AI replaces: it replaces those who decide not to evolve and not to harness its potential. Those who, instead, put it to work alongside people are not replaced. They are amplified.
To explain what we mean, it's best to start with the heart of our craft, creativity, because that's where the misunderstanding is strongest. In the common imagination, creative AI works like this: you press a button, and the work comes out. It has never been like that for us, and not out of artisanal nostalgia, but because of how good work is done. Creativity, when we practice it with these tools, takes the form of a supply chain with two human ends and a technological center. At the beginning, there is a person: the idea, the direction, the why, the decision—often silent—about where we want to go. In the middle, the AI comes in, and here the speed changes nature: from a blank page to the first draft—images, video, audio, text, mockups—takes minutes, not days, and we often link multiple tools so that one ignites the other. Then, at the end, the person returns: the refinement, the check, the judgment on what is worth delivering and what should be left behind.
The machine, in short, is in the middle; people are at the extremes, where decisions are made and where responsibility is taken. For us, AI is a springboard, not the dive; a starting point, not the finished product. It's a distinction that seems small but changes everything because it brings together two things that the debate tends to separate: the power of the tool and the responsibility of the person wielding it. And since it's about power, we add a point that is close to our hearts: such a process should be used wisely, starting with respect for resources and the rights of those who create. Convenience does not erase ethics; if anything, it makes it more urgent.
Beyond creativity, artificial intelligence has entered many aspects of our daily lives with the same spirit. With Claude, we've built a dashboard that monitors customer performance across multiple channels—reach, clicks, conversions, trends—and allows us to detect anomalies as they happen, not at the end of the month. Numbers alone tell us nothing; the work lies in translating them into clear words, and it's precisely in that step that we've gained valuable time. Using the same method, we observe our clients' markets and competitors. Reading what others are doing isn't copying; it's understanding the context and anticipating the direction a sector is moving in. And behind the scenes, AI helps us organize deadlines, priorities, and reminders—not to surveil people, but to protect their time and clarity.
However, if we had to pick just one result, it wouldn't be any of these. The most profound change, one we didn't foresee, is something else: thanks to AI, we've started saying far fewer no's. Anyone working in an agency knows how much those no's weigh. They almost always come in the same way – not enough time, a tight budget, insufficient manpower, an out-of-scale request at the last minute – and these were the situations where, honestly, we had to back down, or accept by lowering our standards. Today, in many of those critical moments, we can respond, “Yes, it can be done.” And do it well, not just somehow.
From the client's perspective, it's an obvious liberation: communication stops being a privilege reserved for convenient projects and becomes available again even when conditions are adverse. But the “yes” has also changed us, perhaps more profoundly. Every time we've been able to accept a challenge we would have previously refused, we've pushed ourselves a little further: we've explored territories we didn't have time to touch, learned new languages, added skills, and become a more complete agency. Every extra ’yes" has also, silently, been an additional piece of growth. We're not talking about quantity for its own sake—producing more has never been the point—but about an expanded realm of possibility, for those who choose us and for those who work with us.
The question remains, which everyone sooner or later asks us with a hint of provocation: if the machine does all this, then what is left for us? The answer, as far as we're concerned, is serene: everything that truly matters is left. Taste remains, meaning the ability to sense if an idea is beautiful, and along with it, what is right, meaning understanding if it is also correct. Judgment remains, which is different from a rule: knowing when a rule is followed and when, with full knowledge, it is broken. Relationships remain, and with them, the art of listening to what a client is not saying. Responsibility remains, which is the ancient act of putting your face on what you sign. And the spark remains – that intuition that arises from no data and that no model, however trained, can produce in our place. AI does not replace any of this. It liberates it.
This is where the math adds up, and we can say it in a simple sentence: thanks to artificial intelligence, we're not working fewer hours, we're working better hours. Mornings are no longer spent downloading reports and pasting numbers; the analysis is ready for review, not for construction. Drafts no longer start from scratch but are already in a good place, and our job becomes communicating an idea as effectively as possible. Project statuses are no longer tracked manually, and creativity is no longer relegated to the end of the day when energy is depleted. We haven't become a smaller, more automated agency, as the common narrative feared; we've become an agency of amplified people. The same minds, the same hands, with multiplied capabilities. More human, not less—because we've given people back the time for the most human part of the job, which is thinking.
So, let's return one last time to the original question. If we absolutely must answer, here's how: yes, AI will replace us, but only if we stop evolving; no, if we choose to use it to liberate people rather than replace them. The truth, however, is that having lived with it for so long, that question no longer interests us. A new, more beautiful, and more difficult one has been born, one we carry with us every morning: what will we do, now that we have more time to think? Demeter is the goddess of the harvest, and the harvest—as we well know—belongs to those who had the patience to sow. The answer to that question, we are already writing. One day, one project, one season at a time.
