Apparently, a new definition for open source AI’s in the process of being created by the Open Source Initiative (OSI), but according to this item, it may be flawed, or outside the purview of that organization. What do you think?
UPDATE
I just saw this item in today’s (Wednesday, October 9, 2024) ZDNet newsletter. It adds new context, and informs that the OSI has released the first candidate for the new OSAI standard.
It says thst for AI software to be open source, the training data must be open as well as the source code…
That makes scientific sense, because all of the intelligence in AI comes from the quality of the training data… the code is trivial by comparison.
That’s an interesting observation, but I’m more interested in the larger question of which open source organization should be responsible for defining open source AI (OSAI). I’m not sure that the OSI has the man power to be responsible for not only defining OSAI, but managing it over time. Personally, I think a new organization should be created, specifically to define, and administer an OSAI standard. What do you think?