[FRIAM] recap on Rosen

[Marcus G. Daniels]

phil henshaw wrote:
> No, that does not work at all.  Patching together a model to suite a symptom
> in retrospect does not help you with being ready for unexpected eventfulness
> in nature that you previously had no idea that you should be looking for.
>   
Never said anything about symptoms.   I did suggest maybe you ought to 
plan on measuring something in particular to see if models (whether your 
own or those you are interpreting) are consistent with reality in a 
statistically meaningful way.  You can posit whatever driving events or 
processes you want in-silco.  A comet striking the earth, people selling 
their organs to increase the profit margins of the companies, the 
importance of prophets in collective decision making, or whatever..