3 ms·
This is not very surprising, actually. Financial contracts & instruments are complex enough - and written in natural languages, often - that it would almost be
by gwern 14y ago
This is not very surprising, actually. Financial contracts & instruments are complex enough - and written in natural languages, often - that it would almost be surprising if you could not construct Boolean circuits or other building blocks of Turing-complete functions. (Have we learned nothing from Wang tiles or Fractran or Turing-complete return-from-libc or the entire area of esoteric languages?) It was already shown that NP problems could be encoded into contracts, so this extension is logical.