Thinking about the OSI "Open Source AI" Definition and how to proceed. Like say they changed their definition to demand that all training data was "available" (right now you only need to describe it) meaning there are URLs that you can access. Think YouTube Videos or social media posts or whatnot. But not all content is under a free license, some explicitly copyrighted with "all rights reserved".
Would you consider a machine learning system trained on that data still "open source" in the intention of the Open Source definition (https://opensource.org/osd)?
@tante no! and it demonstrates the OSI's purpose all along was to water down the radical liberatory potential of free software!
@hipsterelectron I'd argue the same but I'm interested to see how a different reasoning would work
@tante just trying to start a discussion :) i also want to hear that
@tante i have difficulty describing why and how the definition of open source was created to bowdlerize free software (which i also think is insufficient but better)—would like to understand how others relate to the term
@hipsterelectron Yeah. Everyone should hate the OSI because their whole stick is to depoliticize Free Software. But it's there and is used - potentially even in legislation. So it matters.
@tante @hipsterelectron the osi definition is pretty much the same as dfsg, the debian free software guidelines (it is, after all, directly derived from them), and debian looked primarily for a practical way of differentiating licenses.
now the goals of osi were very obviously focused on removing the ideological veneer from the free software (by any name) and make it palatable for the corporate – possibility of which wasn't by any means obvious at the time, but i'd be vary of accusing debian of not being radical enough.
to be frank, i think that complaining that people were not radical enough in the late 1990s and early 2000s is pretty ahistorical, and a very easy thing to say 25 years later.
@mawhrin @tante cc @biocrusoe who has taught me about debian
#Debian classify the kind of #AI systems that the #OSI's #OSAID defines as #opensource, as #ToxicCandy: https://salsa.debian.org/deeplearning-team/ml-policy/-/blob/master/ML-Policy.rst
Without training data you cannot ecercise the freedom to study a #ML system and your ability to modify it is severely limited to fine tuning.
Which is like to say that windows is open source because you can tweak the registry.
For these reasons some open source developers are already moving beyond OSI: https://opensourcedeclaration.org
You are welcome to join!
It's not just matter of securing the #OSD, but to update it with a truly open process, with all developers, artists, musicians, data scientists... who contribute their time and valuable skills to open source: https://opensourcedefinition.org/wip/
