Becoming Protocols

Using AI to express ourselves is powerful — when used with intent

More than 15 years ago, when I started seriously looking into the internet as a source of income, I saw people copying content, rewriting the same thing over and over, optimizing it for search engines, and collecting a few cents per click from Google ads or whatever was the thing at the time. The goal was never to express anything, ideas, nor anyone's self. The goal was to exist where search engines could find you.

I never really engaged in that. It always seemed trashy to me — not illegal, not even necessarily dishonest, but empty. Creating content whose only purpose was to exist, and wasn't even original. In some cases, people weren't even rewriting. They were just copy-pasting.

And the thing is, even when people copy and paste, the content still changes. In Portuguese, we have a saying: "Quem conta um conto, acrescenta um ponto." Whoever tells a story adds a point. Every retelling introduces something new, whether you mean it to or not. A word gets replaced. A sentence gets simplified. A nuance disappears. And that matters to me, because I know how much care goes into choosing words. I sometimes spend more time than I should obsessing over a single word. That word is not accidental. It reflects exactly what I want to say and how I want to say it. When someone replaces it — even with a synonym — it stops being my word.

Because I never engaged in any of that, but also because I thought I had no thoughts worth sharing, I missed out on having a stronger online presence. I still carry the same doubts.

Fast forward to today, and that pattern didn't disappear. It evolved into something far more efficient.

Now we have AI agents automatically going through threads, posting content, replying to discussions, and building visibility on their profiles. Not because they care about the discussion, but because visibility itself has become the resource. Profiles accumulate influence. Influence gets converted into attention, promotion, and persuasion. And at scale, it can shape narratives — or even influence elections, if it comes to that. The internet is filling up with entities that exist primarily to drive engagement.

And here's the uncomfortable part: I'm also using AI.

English is not my native language, and expressing complex thoughts clearly requires disproportionate effort. AI changed that. It helps me structure sentences. It helps me correct syntax. It helps me express what I already think, but in a way that's clearer and more readable. One that I wouldn't be able to output just by myself, not without spending hours upon hours obsessing over it.

But don't get me wrong — it's still my content. It's not LLM-created. The ideas are mine. The intent is mine. The AI just helped me express myself better.

I don't just accept everything it suggests. I go over it more times than I'd like to admit. Sometimes its suggestions are technically correct but emotionally wrong. They don't sound like me. And that matters.

Still, I can't deny it: I write more today because of these tools. They reduce the friction between having a thought and putting it out there. Without them, much of the little that I write would probably remain unwritten.

And that creates a tension I can't fully resolve. I wonder if I'm not also contributing to the unbounded trash being pushed to the internet nowadays. The same expansion of presence for the sake of presence.

The internet was never a perfectly clean and original content place. But with more and more automation — and the quality of that automation increasing — it's getting really hard to tell things apart. Human-created content from just an automated LLM posting about a given subject. The friction that used to exist acted as a signal. Writing something required time and intention. That effort was a form of implicit authenticity.

This isn't limited to writing. In my daily work, I've started using more automation to communicate with my peers — code reviews, feedback, and technical analysis - still my fingers behind it - but I made it a point to keep those things separate. I use a dedicated bot account for automated interactions, and my own account for when it's actually me speaking.

And even when the automation produces a review, I still go back and comment or react to it. Just so people know that I actually looked at their work. That I actually care — not just about the code, but about the work, about them. Because when everything becomes automated, it becomes impossible to know whether anyone actually paid attention.

For me, that distinction matters. It preserves accountability. It preserves presence.

But this isn't free; it takes effort and intent. And I wonder if we're not slowly turning ourselves into something else, hubs in a protocol. Messages go through us, but increasingly they're shaped by the systems around us. We're still there, but our direct presence is getting diluted to a point of almost non-existence.

I don't think the solution is to reject these tools. They're useful. They help people like me overcome barriers that would otherwise keep us from communicating more.

But I think we need to stay intentional about how we use them. We need to remain visible behind the tools. We need to make sure we're still present in what we produce.

Because otherwise, we risk slowly dissolving into the systems we use. Becoming supervisors of automated versions of ourselves, if even that. Becoming nodes in a network rather than people speaking to other people.

The internet is filling with content that exists because it can exist.

Maybe the responsibility now is to create content that exists because we chose to exist in it.

Because we cared enough to choose the words.