Last year, the corporate-dominated web came alive, much thanks to the help of corporate-owned social media platforms - how long will it take until the open sections of the web do so?
And if that has been achieved: with stuff happening in the digital world that doesn’t in the real world, it could actually be worthwile for people to immerse further into the digital world, which up until now was always the problem behind the idea of the “Metaverse” (I’m not necessarily for this; its just something that came to my mind yesterday). Could that be the actual next iteration of the web and realize what was in the past considered the “smart” web or would it be a dystopia?
And one last question that came to my mind: would it be possible to make the LLMs somehow run independently (could the blockchain maybe be finally put to some use here?) and how would all of that be experienced like from the user perspective?
There’s no need to run an LLM on the same system it was trained on. Once the model is built it contains all the information already. If you want a model to live on long term you would just release the file(s) publicly, like hugging face does with theirs, then anyone could use it or host an interface for it.
Is huggingface non-profit by the way?
I don’t believe so. I’m not sure what their long term goals look like.