Proof-of-Work can be Trivial

I'm a firm believer that P equals NP, and thus that proof-of-work can't exist indefinitely.  Phrased
differently, I refuse to believe in trapdoor functions that can't be broken eventually.  Regardless,
I see practical proof-of-work around me and it should be used to one's advantage in protocol design.

A trivial example is requiring all fields in a data structure to be sorted in some way.  This can be
done and checked in linear time and constant space, but is typically only easily checked under these
characteristics.  This thought first occurred to me when I was mulling over the design of something,
noticing I could use much better algorithms, if only the input were sorted.  Mandating normalization
of all things is a step towards removing ambiguity also.  Sorting input likely couldn't be useful as
a decentralized consensus mechanism in any way whatsoever, but it's still proof-of-work, regardless.

I suppose this is proof that work needn't be done more than anything else.

One glaring issue with Internet protocols is how easily they can be centralized by corporations; one
way to fight this is to design protocols in ways that require worse time and space complexity in all
known ways.  Still, it's been shown also that corporations will change protocols to suit themselves.

I suppose this article also ended up trivial.