Cordt

@cordt.zermin.de

I build tools that trust your judgment and write about why that matters. paragraph.xyz/@manual rymi.app

We need a word for an AI thinking that something is expensive and will take a lot of time because it was trained on a corpus where that was still true.

One big advantage of being a builder now (vs an incumbent), is that you can actually build AI first products from the ground up. Instead of a tool with a chatbot slabbed on top, an experience that's actually different.

You need two things to be happy: meaningful relationships and work that provides purpose. People who believe humans don't want to work, learn, or move their body anymore just because an AI/Robot will do the thing in the future, miss the most important part of the equation.

When you find yourself not wanting to do marketing or sales, not willing to speak to the merits of your product, then, maybe, you're actually excited about the technical problem, not so much about putting a solution into people's hands.

This has probably always been true: you need to be obsessed about the problem; only then will you have the endurance to make the solution a success. What changed is that virtually anything you considered tedious, necessary evils, are almost fully solved problems. It's time to be a builder.

Working out products with an LLM has an interesting property: if you force yourself to not suggest any parts of the technical solution and focus entirely on the problem space, you don't accidentally fall in love with any part of the solution.

Ok I'll call it. Fable feels pretty sloppy. No life changing results to be seen, at best a small iteration from Opus, but I don't feel it at all.