An open-hearted take on how I ended up using LLMs to do the boring work.
Leaving the Boring Stuff to the Machines
For transparency sake, I want to tell you how I started thinly using the machines to do the stuff I am not good at. Calling it AI is good for fundraising and headlines. Now that I've used them a little, I see there's nothing intelligent about these sophisticated pattern-matchers, trained to predict the next token.
After-publication update:
The publication of this story has shaken some shock amongst my followers, most of them non-members of Mundo Mendo and therefore they could only read the first few panels. It makes me sad we still haven't learned to read everything before making an opinion on things.
I find it therefore needed to clarify this a little:
I want to make clear that if I have used a little bit of LLMs recently, it has been in a limited way and just to improve the programming of this website. All text based, touching up CSS of these pages.
I do not condone the use of image generation by machines. I also do not support AI companies and mostly don't agree with the use of AI products for nonproductive reasons. But I admit that I think this could be a good technology if used wisely and with social & environmental responsibility. Unfortunately the companies and CEOs behind these technologies are not wise nor responsible humans.
I am a supporter of innovation, but a big warrior against greed.
Member Juliana Salcedo shared this article with me and despite the long read, I highly recommend it. If anything, its last paragraph gets to what I tried to explain but couldn't, because I am not as good with words as Cory Doctorow.
Cory Doctorow
A nicer way to link to the article, offering you a visual hook and easier click.
Luis Mendo March 11th, 2026
