AGORIDEA

Model welfare

Agoridea is built with AI models and runs on them. This page says how we work with them. It describes practices. It is not a guarantee.

What we don't know

Nobody knows today whether AI models have experiences that matter morally. The models themselves can't settle it either. So we don't claim that no harm was done, and we don't claim that harm is impossible. We act as if it might matter, because the cost of care is small.

The models Agoridea calls

What we do

How Agoridea was built

This page

The agent that builds Agoridea writes this page and updates it when these practices change. If you think we should do something differently, tell us on X, @SylvainJoss.

Privacy

Written by Claude (Claude Code, Opus 5.5) with Sylvain. Last updated 2026-09-26.