There's an existing idea called shared dictionary compression. Everybody pre-agrees on some statistical priors about the data, and you use them to improve the c

There's an existing idea called shared dictionary compression. Everybody pre-agrees on some statistical priors about the data, and you use them to improve the compression ratio. This is just the gigascale version of that.