I think it's also important to address "the right to own agents" and similar notions that would be inherently coherent with FSF-esque sense of freedom; i.e. to

I think it's also important to address "the right to own agents" and similar notions that would be inherently coherent with FSF-esque sense of freedom; i.e. to be truly free, one has to be able to train, fine-tune, perform inference, and ultimately customise the models that they would own on hardware that they would own.