Nothing in software is ever "the future." How many things change as fast as software technology? For many of us not concerned with "building our brands" through "resume driven development", but instead focused on enabling rapid and reliable production deployments containers and orchestration have improved our results. Obviously this will not remain the dominant paradigm forever. The next thing may be an evolution of it, or it may be a take in a different direction that leaves kubernetes a dusty and unused side-branch in history. Wouldn't be the first time. Regardless of how likely that is, or what the drivers of the eventual result are, I am not sure what the author wishes people to have done instead of employing containers? Should we have just kept on installing dependencies and services onto VMs while awaiting the one true new thing that would last forever?