In one of the first studies of the disciple of Human Geography into distributed shared ledgers, Blankenship \cite{blankenship2018} conceives blockchains as production spaces where developers are the dominant class within the social and technical spaces of the technology, have ultimately leveraged their knowledge and power dynamics to accumulate wealth via the token value, and then shifted into the role of investor. This necessarily involves automation (exploitation of automated robot labor) and obfuscation of the mechanisms of production- geographic borders are defined via conflicting abstract conditions (social, political, and economic), and put within the qualitative context of social dynamics.Humans not only trust in the source, they trust the structure -you generally do not care about who wrote a diet article (even if a change in lifestyle can have lasting impact on health) as long as "structure" suggest the writer is not a charlatan. A similar behavior is observed in crypto markets, where traders and investors keep lists of Twitter accounts that they trust to rely accurate information about the state of the market, and that are facilitated by other traders: it does not count only who is saying it, but who is following -this is part of the social fabric of crypto markets, the structure of the network encodes tacit knowledge and reflects abstract conditions and boundaries. The distance trust metrics have very tangible implications for individual and corporate purposes; "a member of my group said " (even if he had materially different attributes) is generally better than what an outsider says. The implications in terms of the theory of the firm (Coase): you do business in the proximity of your circle (your trust space) where trust is secured; even if it is more expensive to produce in your inner circle, and it is cheaper to acquire in the boundary (e.g. potential partners) -but going beyond that will require a significative leap of faith, and the associated risk should be priced-in. The AI will understand this human bias (as shown in Figure 2), as a trust differential in terms of metric entropy.