Re:Show HN: Cracking passwords with a simple genetic algorithm
by eternalban on Monday February 29, 2016 @09:51AM
(#11195725)
There are interesting developments in formally defined and verified programs (including implementations). For certain domains that are non-intuitive, exploring the solution-space is now much nearer in reach.
For example (as I noted in private communication to John Koza) exploring the protocol-space of distributed consensus algorithms on top of something like Microsoft Research's IronFleet (with the GP generating Gafny code) appears to be both possible and imo a fruitful avenue for researchers in the field.
[p.s. Note that Genetic Programming is distinct from Genetic Algorithms.]
For example (as I noted in private communication to John Koza) exploring the protocol-space of distributed consensus algorithms on top of something like Microsoft Research's IronFleet (with the GP generating Gafny code) appears to be both possible and imo a fruitful avenue for researchers in the field.
[p.s. Note that Genetic Programming is distinct from Genetic Algorithms.]