Jason Antic

1/ My unpopular take: Free and Open Source Software culture often goes way past the point of absurdity in its idealism. It often winds up doing to software what file sharing did to music: It devalues the work and the creators behind it, and it breeds an entitlement culture.

File sharing was just about dissemination but open source is about collective creation so I don't think the analogy is useful. I understand your point, that the pressure to make cheap sw lowers quality unless userbase is huge.

Collective creation is the ideal that is promoted around open source. But I think the actual usage vs contribution ratio is tilted overwhelmingly heavily towards non-contribution in practice. And what really gets emphasized and expected is the "free" part.

So you don't think that the deep learning libraries/models you rely on have benefitted from open source? All the model architectures that are available, crowd sourced bug detection, user driven issue raising, modularity thanks to transparency...