With situated software, you're in a direct relationship with the software.

It's not some black box built by someone else.

It's software that you've evolved and grown yourself.

You've been playing with it, developing your own "theory of mind" to predict what it will do, changed it to behave closer to what you want.

Situated software is active software, not passive software.

You're in the loop with the software.

This solves a confusing riddle:

Why is ChatGPT so compelling, but most LLM-powered UIs are underwhelming or hard to trust?

In a ChatGPT conversation, you're in the loop with the LLM; you can see how it's responding and tweak your interaction.

In traditional UIs you're on the receiving end of a final prediction with the LLM somewhere deep inside, impossible to access.