I agree, good comment. There is so much big tech in foss/oss community, like you gave example with repos and with react. People working for free for big tech. Getting something to put in their CV or just bootlickers.

Every open source or foss from company is in company interest, it will never be in community interest.

Nobody is paying for web frameworks like React so get a bunch of Meta bootlicking nerds to improve it for them for free.

React was developed internally by Facebook, then open sourced. While a big part of this is getting regression testing and bug fixes for "free," the primary motivation was to outsource training costs. By making React publicly available, Facebook no longer had to train new hires how to use it. Instead they've created an entire industry of third party code bootcamps and a long line of applicants climbing over each other who are already familiar with their tooling. It was a move driven by the objective of undermining labor power within the firm, to make their software developers more fungible and replaceable. Though this strategy has been replicated elsewhere, it was somewhat ahead of its time, going beyond the typical approach of just using "open source" software because other people will maintain it for "free."

The Open Source Initiative has always been like this. It has ALWAYS been about producing higher quality software and the superiority of its development model. It is designed not to challenge the existing status quo. That's their preogative.

This is why the term "FOSS" is harmful. It attempts to generalize groups of people in vastly different stages with vastly different goals.

Reject FOSS, one should clearly state what they mean and want. We are united in varying degrees of struggle against proprietary software, but each of our goals and circumstances are different.

FOSS is so popular because it homogenizes us into a box that can be categorized and compartmentalized. Labels are comforting, but they can also be harmful. One thing is for sure, "FOSS" isn't a threat to these corporations.

Meta and Google open sources PyTorch and Tensorflow so people can hopefully make one better than the other.

That's a bit optimistic at this point... PyTorch is basically the Windows of machine learning libraries -- it's not particularly great on its own merits, because a lot of core features (XLA, jit compiling) are clearly added as afterthoughts and have a lot of very apparent issues, but everyone uses it because everyone uses it. It is a perfect expression of why maybe "move fast and break things" isn't such a great philosophy for important libraries.