A piece of this is to eliminate the friction of legal text by open sourcing legal templates, in a graph, on GitHub. The key word here is "source." Legal documents currently are blobs, not source. See github.com/CommonAccord/ Lots of exampless, including the YC SAFES, Series Seed, GA4GH data sharing, etc. It's not full algorithmic, but compatible with algorithmic approaches, like Mike Genesereth's and Meng's.