> Given a binary choice of a developer not releasing any code at all I mean it wasn't binary earlier, it was "to get more model developers to release", so not

> Given a binary choice of a developer not releasing any code at all I mean it wasn't binary earlier, it was "to get more model developers to release", so not a binary choice, but a gradient I suppose. Would you still make the same call for software as you do for ML models and weights?