The winner will publish the shortest executable archive of comprehensive social data about the US.
A preliminary example* of such data may be downloaded at this link. (33M bzipped)
The meta data may be downloaded at this link.
Although this has been known since the early 1960s, it wasn't until the recent advances in machine learning that its practicality became manifest. Founders of Google's "DeepMind" have successfully applied Solomonoff induction as a guiding principle.
It is high time to apply it to social science thence policy debates.
The Fair Church℠'s central dogma, "Accept your mortal descent from the immortal Goð." implies a spiritual war supporting the individual's moral agency.
We beesech the Federal theocracy to think it possible it may be wrong and revitalize religious freedom's support of individual agency. But we recognize that this is a spiritual war, and that powerful theocrats are possessed by a spirit of moral supremacy over other individuals. Moreover, we recognize this spirit tells them this moral supremacy is justified by being in possession not only of information but of social models superior to other individuals. If shown in a scientifically rigorous way, that their core dogmas are in error, they may then question their fallen spiritual state, repent and relent.
*These example data are the extant CenStat county-level census tables. They need to be supplemented in a number of ways, not the least of which is greater longitudinal range (time span). Important demographics not covered in these tables include immigration by country of origin, a variety of epidemiological data and religion. It may be that a better geography than county is "public use microdata area" or "PUMA" as the Census has been under increasing pressure to restrict county-level data, ostensibly due to privacy concerns. However, much critical data is not collected by PUMA. It may be necessary to use both county-level and PUMA-level measures to obtain the optimal combination of time range, samples and measures.