Yes, this is deep and complex.
Yes we are dealing with deeply nested sets of complex adaptive systems (up to 20 and sometimes more levels).
The classical notion of “Truth” (that one can be absolutely certain of something) must be one of the first casualties, and it needs to be replaced with contextually relevant confidence. And in some contexts that confidence can be very high indeed, within 20 decimals of unity in the case of modern processors with regulated temperature and power supplies.
In some contexts, some systems can be very reliable indeed, in other contexts not so much.
Of course some things are socially constructed, but not all things.
Of course social construction has a role in our perceptions, but so do evolved biochemical and social constructs, and choice.
The issue is not with the claim that some things are socially constructed, that is obvious.
The issue is with the claim that all things are socially constructed, and that is not a simple claim to disprove, and I am satisfied that it has been disproved beyond any shadow of reasonable doubt.
So where does that leave us?
We are in a deeply dangerous border area, between scarcity and abundance.
We have a socially dominant valuation mechanism (markets) that are scarcity based and cannot give a positive value to the universal abundance of anything.
We have technologies capable of the delivery of universal abundance of a large and exponentially growing set of goods and services.
Most people are still firmly inside of the accepted social construct of money and markets, and cannot yet step beyond it to see the existential level dangers present.
And certainly, yes, markets do perform many levels of very complex and necessary risk management, including distributed cognition and distributed risk management and distributed governance. So developing effective transition strategies is a seriously non-trivial complex task; and it is one we need to complete in the next decade.