I would interpret an ab initio, or from first principles, method as describing something from the most generic sets of principles that are available. What that means, may be time or field specific. In most of current engineering, and for a long time in history, classical mechanics is/was this basic set of principles and laws. However, today it can also be considered a special case or subset. Perhaps one could say that a current ab initio method would have to be based on the theory that currently holds the most “explanation power”. From this perspective, it makes sense to call quantum physics an ab initio model.