The Daring of Ignorance: When Knowledge Blocks Software

Since Antón was Crema and, of course, now that he is C. Tangana, I have closely followed the career of an artist who never stops unsettling the status quo. First with rap, then through that transformation where he blended rhythms and influences that no one saw coming in Spanish urban music.

Pucho is an inspiration: someone who, as he says himself, "without singing or staying in tune," has gotten an entire country to listen to an album like El Madrileño in an era where everything seemed destined to mimic the US sound. But I’m not here to promote him (though if you haven’t heard it, you’re missing out). I’m here to talk about a talk he gave recently at the MOP Foundation with Santos Bacana about his creative process, and how his words blew my mind when applied to the world of digital products and AI.

1. The Daring of Ignorance

Antón mentions that ignorance is, paradoxically, one of an artist's most sophisticated tools. When jumping into directing a film or composing with flamenco masters without being an expert in harmony or a cinephile, what we see is someone who isn't afraid of what they will find.

In the world of software, the opposite happens. When you try to build something new with AI, the mantras of the technical (or "engineering," which sounds more real) establishment always appear:

• "The code is bad."

• "It won't be maintainable."

• "We are generating technical debt from minute one."

• "We depend too much on the model."

Honestly: all of that already happens with the current production model. The difference is that now, the fear of "doing it wrong" blocks us. The key is to take the plunge, be curious, and if it falls to pieces, let it fall. Code that breaks but exists is preferable to a perfect architecture that never reaches production.

2. The Danger of "Doing it Right"

When you know exactly how something is done, your brain seeks the path of least resistance: what has already been proven. Knowledge gives you security, but it also boxes you in.

As Pucho says, acting from a naive or wild place allows the error to become a new aesthetic proposal. This might be the key to the future of software: a stage where "one size fits all" (the standard) is worth less and radical personalization—even if the underlying code is "ugly"—becomes the norm.

3. AI and the "Dirty Workshop"

Antón describes his studio as a workshop: a dirty place, with worn tools and unfinished things. Genius appears in friction, not in efficiency.

I imagine a future where AI isn't a perfect assembly line, but that workshop tool that allows us to "feel our way in the dark" faster. Pucho says he needs 9 hours of studio time "thrown in the trash" so that in the 10th hour, the hit appears. AI is a machine for throwing those 9 hours in the trash in 2 seconds, allowing us to reach the breakthrough sooner.

4. The MVP: A Necessary Mess?

Here we connect with the concept we work on at Fika. The Minimum Viable Product is the equivalent of Pucho's dirty workshop. Projects you launch with the purpose of learning and testing that side project in the market.

• The Expert Trap: An engineer hates the MVP because they see the seams and the code. They suffer.

• The Creator's Reality: If you wait for the code to be a work of art, the trend has already passed. As Santos Bacana says: "If it's bad, no one will ever know."

The MVP is your safety net. If it has "duende" (utility or emotion), the user will ignore that the code is ugly. What matters is the impact and what it produces for the customer, not how it was built.

5. Conscious "Fake it till you make it"

There is a difference between being ignorant and choosing the position of the ignorant. Antón knew that urban music was going to explode and decided to "occupy the space." He didn't wait to have the title of "King of Spanish Pop"; he acted like it until the rest of the world agreed.

Success is an anomaly within a routine of failed attempts. It isn't divine enlightenment at minute one; it is the result of enduring the frustration of "fumbling in the dark" until the thread appears. To get to project number 11 (the one that succeeds), you had to go through those previous 10 hours of trash.

Conclusion: Knowledge as a Filter, Not a Brake

Pucho gives the example of classical guitarists who said flamenco players played "badly" because they didn't follow the rules. In AI development, many will say your way of building is "crap" because it doesn't follow classical patterns.

But if that code supports a disruptive idea that reaches the user (just as El Madrileño reached the people), then what was "done badly" becomes "done rightly."

The modern developer must stop being the person who writes every line and become the one who feels the system, detecting where there is a hallucination and where there is an opportunity. Moving from being the executor to being the curator.

Do you identify with that feeling of frustration when seeing code that works but is "wrong," or do you think we are entering an era where the "how it was made" matters less than "what it achieves"?

Success is not divine enlightenment at minute one. It is the result of enduring the boredom and the frustration of "fumbling in the dark" until, by exhaustion or by chance, the thread to pull appears. This is what AI allows... understanding which thread to work with.

And in the end, this is the key: Understanding that this is a process and that often, to get to that 11th project and have it succeed, you had to go through those 10 hours thrown to the wind—which, thanks to today's tools... you can now navigate more consistently.

Thanks for reading me.

Gonz.

¿Tienes ganas de más? Visita itwasgonz.com para conocerme mejor.