Structure is opportunistic in the sense that it results from energy bleeding out from the big bang into other fields, not only quantum fields but things like position - thus we get large scale isotropy.
I followed Shrodinger's ethos to its limits and all I got was perfectly isotropic relativistic spacetime foam.
I suppose that's what our universe will eventually decay into - and then what happens is up to anyones interpretation (conformal cyclic cosmology or what have you..)
I need to implement gravity, some type of coupling that can prevent equipartition of states from happening so quickly.
In any case, nothing interesting happens in small universes by the process of evolution. Evolution requires a balance between specifity of search/selectivity and magnitude. Less specificity will require more brute force magnitude. Just like any search of a space or data struct. Silicon just doesn't have the magnitude that even a mole of molecules does. Not even in the same ballpark.
The source is up here and easily portable to platform other than Windows
https://github.com/churchofthought/ScatterLife
I need to update the video - the video is before I added a projection matrix to fix the hexagonal skew.
Another framework but using complex valued cells:
https://github.com/churchofthought/HexagonalComplexAutomata