[dumbfoundded]

I don’t understand how this model can ever be predictive. It seems like a good way perhaps to create an approximation engine but I’m not sure what sort of predictive insights you can gain from this approach.

In order to be useful, this model should either create a new testable prediction or speed up computations in existing models while retaining accuracy. It seems to be in the latter camp. I would like to understand more about why this model is more computationally efficient.

Perhaps there’s more work to be done on the process of generating rules or limiting the types of rules. Arbitrarily choosing rules to create properties that look similar to observed physical properties doesn’t seem to point to a fundamental theory.

[acjohnson55]

I think Wolfram makes the point himself that the model may not directly be predictive in the way that states "encrypt" their predecessors. I guess the best bet is to show that there is a direct connection between his model and higher models, because then you can start to look for physical manifestations of things his lower-level model predicts that don't exist in higher-level models. Kind of like how GR keeps getting reinforced by verifying its predictions with phenomena that hadn't been considered before its advent.