How I use AI
As of today, August 5, 2026, here is how I use generative AI/LLM tools.
I have a Claude paid subscription, and I use it to direct my questions, either on my mobile or through their website. I think of Claude the same as I think of any other AI: as my thinking partner.
I think of personal websites like my own, as the name "personal" already implies, as personal attempts at expressing myself.
I once read Cris Dias asking about the meaning of the word essay (ensaio, in Portuguese). It comes from the French essais, or tries (tentativas in Portuguese). It was Michel de Montaigne who popularized the term, around 1570. Cris says that's the reason we use this same word, in Brazil, for the theater: “that's not quite ready yet, ok?”.
I declare, therefore, that anything written here, be it a blog post, be it a note shared in my digital garden, is an essay. A big "Here's the thing, I've been thinking that...". Holding my personal take, my personal view, my personal interpretation on things — never those of Claude, or any other AI.
Ethan Mollick proposes a way to decide when AI is worth using: not "is AI better than a human", but "is the best AI available right now better than the best human I can actually reach, for this specific problem, in this specific moment".
I find Mollick's framing very honest, as the real comparison is never against some idealized expert, but against who's actually there. I mean, I don't have an expert on call at 11pm, when I have an idea coming out of the blue for a note, that I want to dive into and explore. That's when Claude comes in: not as a replacement for the best available human, but as the best available human I can't otherwise reach. While I write, I would ask this partner questions such as "how many flags are there in the world that are not shaped as a rectangle?", or "tell me about how an astronaut sleeps during space missions".
As for programming: I have been working with either information technology or information technology related stuff for more than two decades, managing IT projects, writing and reviewing business cases and dealing with brainstorming and designing processes. But I don’t write code by hand anymore (except for the occasional VBA macro to automate what not). The logic stayed, though, and that’s what I lean on when I decide to vibe code. Yet, I don't want to sound like I'm vibe coding all the time (because I'm not). And to be clear: this personal website is about essays, prose and ideas — not about programming.
If this changes in the future, I’ll update this page at this permanent URL: https://danielsantos.org/ai
↻ This page was last edited 3 minutes ago