Being a bit hand-wavy with it: It’s akin to torrenting music/movies. The torrented files are lossy compressed representations of the original waveform from the music producer. Limewire, or Pirate Bay, or whatever provide interface to retrieve them (download or stream). The model weights are a form of lossy compression, and inference is like a document retrieval.
One may say, “it’s like an employee working at company X, then going to work at company Y, they retain their knowledge and experience.” I would say it’s more like, employee going from X to Y, but retaining audio and video recordings of all interactions he had, notes, documents, and other proprietary info and bringing it to company Y.
This concept has specific technical meaning -
https://www.nolo.com/legal-encyclopedia/fair-use-what-transf...
It seems obvious to me that to call model weights "lossy compression" is not only incorrect from a technical (software dev) point of view, but also from this legal perspective.
The weights serve a different purpose than the original works from which they are derived, and wouldn't/couldn't POSSIBLY exist were it not for the original work of the authors of the models.
It's bad practice to go around espousing strong and condemnatory opinions about topics you don't have a full grasp of. In this case, it's both the technical details and the legal system.
It makes you look like a fool and costs you your credibility amongst peers in future encounters.
I disagree, the internet wouldn't be half as full of knowledge as it is if it weren't for the loudly ignorant giving the experts somebody to correct.