I am aware of the fact that bitcoin ASICs are a pure hardware implementation of something similar to this pseudocode (only in transistor logic that can't be intercepted anywhere):
while(nonce <= 0xFFFFFFFF){
if (SHA256(SHA256(hex2bin(Header..0x00000000..nonce))) < target) return nonce;
++nonce;}
I am also aware that this particular series of computational steps is really only useful for performing this specific proof of work.
What I am suggesting is that someone will eventually devise a way to design a particular service or product that specifically uses the bitcoin-proof-of-work algorithm, in order to monetize the abundance of no-longer profitable ASIC devices that are now taking up space but people will be unwilling to get rid of due to how much they originally cost.
We already see this with SHA256 Alt coins, which I'm certain will be popping up in increasingly ridiculous numbers very soon.
Think out of the box a little bit. Let's DESIGN something that can use this particular hardware function loop as-is.
We can use them to collectively find SHA256 collisions. Or as a living proof that SHA256 is safe.
But the economics of this is failed: If we ever find one collision, the BTC/USD price could drop and many ASIC owners are long BTC.