@b >all their technologies are open source
What are you trying to say here?

"open source" refers to specific licenses for software that meet 10 requirements: https://opensource.org/osd (yes, it's clownflared).

A lot of the ((((AI)))) technologies are source available, but are proprietary software using a license that doesn't even qualify for "open source" (you would think they would at least craft a proprietary software license manages to pass the 10 requirements).


Current ((((AI)))) datasets seem to be built up of mostly copyrighted data that's licensed under the default "All Rights Reserved" terms, except that data happens to be publicly available.

Lots of companies are willing to claim that they "own" many outputs, even though that output is usually a mashup of many copyrighted works under incompatible terms (as different "All Rights Reserved" works usually can't be combined without permission from each author).