AI ethics – Underlying compute power
Yesterday I turned on my LG smart TV and to my surprise I found I couldn't continue using some/most of the features unless I agreed to the new T&Cs that had been rolled out overnight to my TV.
For the first time in my life I read the T&Cs about privacy. In order to continue I needed to agree to my personal data being shared out side of the EU.
My immediate thought was WOW my TV overnight became the TV from 1984 (to my knowledge it might have already been for a while). So when did my TV start watching(listening) to me?
So in the shower I started to think about what could have a big impact that all us where we wake up overnight to find new T&Cs that significantly affects us. I kept thinking about AI and how pretty much all AI models are trained on NVIDIA hardware.
If NVIDIA decided to push a T&C update that said it was against their wishes to train AI for XYZ it would significantly affect the future.
This leaves me with some thoughts. - Why isn't there significant competition in this space like AMD? - Why doesn't AI ethical groups push for decentralisation for compute power? - Should someone like an AI developer avoid using NVIDIA? - What are the alternatives? - Does this even matter?
I'm pretty much a noob for this sort of thing but I tried to do some research on this before posting but google just gave me clickbaity articles so I would much appreciate your brainpower on this.