Self-Improving Software

3 iFelix 2 7/21/2025, 3:04:26 PM fcf78023.twenty-dev.pages.dev ↗

Comments (2)

iFelix · 8h ago
Hello HN!

There’s a lot of focus (and money) on scaling LLMs, but while building Twenty [1] we’ve observed that current models are already capable enough for most use-cases. We think that what’s missing today is a new kind of software architecture. One that evolves, learns from feedback, and compile complex patterns to adapt to users. I wrote about this and some of our learnings in the article above, curious to hear your thoughts!

[1] https://github.com/twentyhq/twenty

danoandco · 6h ago
reminds me of the "memory for agents is a moat" but for software

memory for software