I have a theory about NFTs.
One of the issues with running any system on the internet where users can upload files is that you have to store those files and make them accessible to people.
Which is fine. But often times users will upload the same file repeatedly or with only minor trivial alterations and that is how you end up with a multi petabyte server farm completely full of duplicate memes and reposted video content.
If there were a system that could uniquely identify each and every single file that is uploaded to it and then on the fly cross compare that file with previous uploads and identify when the file is a duplicate, that would be very handy.
If such a system like that existed, then system would only need a single copy of the file to share out to every single person that wants to view it, saving possibly hundreds of gigabytes of storage space.
Something like NFTs could be used to accomplish that, if it were fully fleshed out and developed.
It would probably need one of the new neural network systems overlaid on top of it to analyze the content and cross compare it, and to single out minor differences and create overlays to account for them.
For instance, if you have a base image on knowyourmeme.com, then all the AI would have to do is separate a layer that contains the text in text format with the information of the font that is displayed on top of the image, and it could share up millions of variations on a base level meme for only a few megabytes of storage space for the text information.
If your computer system already has the base meme file on it rather than sharing the entire file to you it could just share the text overlay to your computer and decrease transmission file sizes and time as well.