[Osmose]

LLMs are not a general public benefit. Artists whose work is trained upon by text-to-image models aren't made any more whole just because Meta has to share its weights—it just means it's even cheaper for the folks impersonating them or effortlessly ripping off their style to keep doing so.

Meta really does not need to be subsidized when they have so many resources at hand—if LLMs are really hard to train without that much data, then perhaps that's a flaw with the approach instead of something the world has to accommodate.