Armstrong, even with his supercomputer, his presumably bright team and all of his work cannot model things for which there IS NO DATA (some Black Swans do not equal other Black Swans, also note that Taleb defined a Black Swan as an UNKNOWABLE & UNPREDICTABLE event, among other things).
"The butterfly flapping its wings in Beijing may affect grain trading in Chicago..." <=== You can't model that!
There is nothing unpredictable. The entropy of the universe is not infinite unless you consider infinite time into the future. So here and now, there is nothing unpredictable. It is just a question of is your data set comprehensive enough.
There is a hidden order actually even to Butterfly events.
There are some new data that Armstrong can't see. He can see that anonymous crypto can be a frontier. His computer model predicts frontiers to arise (even MA has mentioned grassroots creative destruction), but he personally can't see it because anonymous crypto was never created before.
Entropy is increasing, but it is doing so in an ordered way because again entropy is not infinite. It is the small things and the short-term that can not be predicted accurately because the new entropy is seeping in there. The big picture (macroscopic) entropy morphs less radically than microscopic entropy on each cycle.
Time may just be an illusion but it is the illusion we are assuming here for our purposes.
No risk no reward. But if you understand the unstoppable trend of technology, it doesn't look conceptually risky. Mankind can not turn back the clock on technology. The big risk is on implementation, human failings, misconstrued marketing, etc..