Should we enable public binding precommitments?

The abil­ity to make ar­bi­trary pub­lic bind­ing pre­com­mit­ments seems like a pow­er­ful tool for solv­ing co­or­di­na­tion prob­lems.

We’d like to be able to com­mit to co­op­er­at­ing with any­one who will co­op­er­ate with us, as in the open-source pris­oner’s dilemma (al­though this sim­ple case is still an open prob­lem, AFAIK). But we should be able to do this piece-meal.

It seems like we are mov­ing in this di­rec­tion, with things like Etherium that en­able smart con­tracts. Tech­nol­ogy should en­able us to en­force more real-world pre­com­mit­ments, since we’ll be able to more eas­ily mon­i­tor and make pub­lic our pri­vate data.

Op­ti­misti­cally, I think this could al­low us to solve co­or­di­na­tion is­sues ro­bustly enough to have a very low prob­a­bil­ity of any in­di­vi­d­ual ac­tor mak­ing an un­safe AI. This would re­quire a lot of peo­ple to make the right kind of pre­com­mit­ments.

I’m guesing there are a lot of po­ten­tial down­sides and ways it could go wrong, which y’all might want to point out.