He believes that migrating to new hashing algorithms as a policy would significantly increase decentralization and hence security. The impact on existing miners could be made pleasant by gradually moving the block reward from the previous hash to the next, such that both are accepted with different rewards. An appropriate rate could possibly be calculated from the difficulty. The author suggests developing the frequency of introduction of new hashes such that once present-day ASICs are effectively obsolete anyway due to competition, new ones do not have time to develop. Karl Semich is interested in hearing thoughts and concerns regarding this matter.