Reflections on AI

Reflections on AI

April 7, 2026

Lately I’ve been thinking about AI. These days, everywhere you go, you’ll always find something like: “Use this prompt to generate that,” “upload your résumé for this,” “give access to the newest agent that will automate everything.” It can send emails for you, manage your life, automate your job search for you. All of this brings an enormous amount of convenience to our daily lives and helps us optimize our time. I personally use AI for everything: studying programming, at work, learning languages, and so on.

But a question came to me: how far can this take us? What kind of consequences might it bring?

This article is more of a musing from someone who admires AI, but questions a possible problem in the future. Today, what we see is everyone benefiting from it, and if you ask anyone about AI, they’ll show you N reasons why AI is the future, has changed the world, and will keep changing it.

The problem I see is that no one talks about the burden; we all only talk about the bonus when it comes to AI, and that worries me. Because, as an architect, I keep one thing in mind: every technology I implement, I have to know its trade-offs. If I don’t know its trade-offs, I’m not ready to implement that technology yet.

My concern is this centralization of data from all over the world. People send their photos to improve them, people craft prompts to create tests or systems, people send their résumés, people send exam results for the AI to read. With all this information, patterns of behavior, health patterns, and many other patterns are being generated.

With the algorithms that help us, I think, the same algorithm can also harm us. One example that could arise is the denial of a bank loan: with patterns we send — photos to AI, searches for flights with AI, recommendations of places to visit via AI — all of this could be extremely valuable data for banks. They could generate patterns suggesting that person might be someone who won’t repay the loan because they spend a lot of money on things, making them a potential defaulter.

An insurance company could deny a life insurance policy to someone because, in their medical records — exams we send to the AI to read — it creates patterns showing that person has a higher chance of developing a certain disease or a much shorter life expectancy.

That’s just speaking about big techs and other companies that can and will benefit from this. Governments around the world can also leverage this information. And we all know that history shows us that governments have always accumulated information about their citizens to better understand and fine-tune the state machinery: in some cases, to improve a given region; in most cases, using this information to persecute their opponents. This is all in a world without AI, where governments and companies needed more human work to do that data analysis. Now, with AI, which can analyze data without rest, finding patterns. What does the future hold for us?

This is just a reflection I’d like to share about the future of AI. Are we exposing ourselves too much? Could the solution be using tools that claim to defend the privacy of our data?

Perhaps the greatest risk of AI isn’t what it does for us, but what it comes to know about us.

💬 Comentários