Well I kind of thought of this few months ago. And I never thought somebody is thinking deeper about this. I believe this is really possible. And this will not take so much work. Just like the primes or any algorithm or solution set, it is possible to make this happen.
So what kind of team are you looking at? I guess if you needed some more help, contact the developers and not in this forum. I guess some will bad mouth your posts and troll this thread.
This is possible. It can be done. I would have tried to keep this a secret and work on it on my own, but I'm 42 now, and my brain pan just isn't what it used to be, I can't seem to learn new things very fast, and I can't imagine how to program something like this, because it seems too hard to program. But the actual theory itself IS easy.
The encoding portion can be done on the fly it's so fast. So we can get rid of all of those min-harddrives clogging up space in our smart phones (well, we can install apps on them of course, but for saving any other files, they can all be encrypted until its time to view them, when they are moved onto a standard drive for viewing and then deleted once they've been watched, keeping the crypto code in the vault for safekeeping permanently.
We can then offer all cameras, camera phones, video cameras, etc ... all medias that encode data into binary data, the ability to record video in chunks of say 500 Megabytes at a time, each with its own code that is only a FEW Kilobyte's each. When you want to watch the video, it gets moved and decoded onto the standard hard drive for watching, and then dumped upon completion.
Imagine where we could take this theory if it indeed becomes a reality:
1) No more SD cards for storing large videos. Recording video on the fly straight into Crypto Key.
2) Facebook and other such companies who do backups could stop having to buy huge server farms that are larger than we can comprehend to store their data. Back ups would also store very fast, since the encoding portion is nearly instantaneous. The decoding part is much slower though. It requires brute force searching through branching timelines of huge data sets to find one "solution" (a needle in a haystack) and that's the only place this theory has its drawbacks. In the decompression part. Until we programmed it (and perhaps we could even program it to use Asics instead of GPU's) and saw how fast it could do such kinds of hunts, it's hard to say if it's way too slow to use. That's my one fear. We'd have to see.
3) No one could sniff your data enroute and see what it is as it's being transfered.
Thanks for your positive comments, you guys. I appreciate it.