Post

@adamhjk.me thinks that we should just let the LLMs go wild, but @matthewsanabria.com sees the value in continuing to do tasks ourselves that LLMs might automate. While those with decades of experience benefit from the speed, those with less experience benefit from the friction and the struggle.

1 repost0 likes

I’m not “submitting to the LLM” - I’m building the machine that builds the machine. It’s not a question of shutting your brain off and ignoring quality - that, of course, builds awful things. There is of course value in understanding how things work, of how structure and architecture

Is expressed in the code itself. Instead it’s a question of how do you build a system that continuously improves on the outcomes? Nobody, including me, knows yet what it means to teach people to build in this way. Literally nobody has actually been taught!

But the idea that there is no learning there, no loop to explore, no knowledge of the system - I think that’s not going to age well. There is *plenty* to be learned. And the “domain” in domain driven design isn’t *software* - it’s the domain of the problem!

Which routes us back to what kind of optimization are we really performing? My business partner can’t program - but we built a financial modeling app that replaces the spreadsheet we built over 15 years in a few days, and it encapsulates decades of our expertise.