@ben #Funfact: all "#learning" is #FairUse, otherwise you'd be a perpetual #DebtPeon to ]whoever made your schoolbooks and created whatever media you ever consumed](
http://felixreda.eu/2021/07/github-copilot-is-not-infringing-your-copyright/ ) !

GitHub Copilot is not infringing your copyright

Felix Reda

@kkarhan @ben
While I agree with the proposition that "#learning" falls under "FairUse", I have some lingering doubts:

- Can we truly consider machine learning as a form of learning?
- Is it valid to compare the training process of a machine with human learning?
- Considering that copyright is rooted in intellectual creation and the intention to create, it's worth questioning whether machines, devoid of intellectual concerns and creative intent, can be held accountable for copyright infringement. Perhaps the responsibility lies with those who train the AI, ensuring compliance with the copyright licenses of the input data used.

@char @ben as shit as it sounds, "#learning" isn't something unique to humans or mammals.

If you have read the article linked you would've seen that trying to "curb in" and crimimalize #MachineLearning would basically shoot research and studying in the foot - with a fully loaded stechkin set on fully-automatic - and then wounder why there's blood on the floor...

@kkarhan @ben
I actually read the linked article, before my previous toot; don't agree 100% with it.

@char @ben neither do I, but I also think #Copyright and #IP nowadays is more of a weapon for IP hoarders than a social contract to reward innovation.

Copyright Forever Less One Day (REUPLOAD)

YouTube
@kkarhan @char @ben There's a big difference between "learning isn't unique to mammals" and "learning is what this particular set of algorithms is doing." For me there are just too many substantive differences between gradient descent and what happens in real brains to call the former "learning".