> systems that are many times smaller (maybe many orders of magnitude smaller) than today's biggest systems. This is a dangerous and misleading metric. You're

> systems that are many times smaller (maybe many orders of magnitude smaller) than today's biggest systems. This is a dangerous and misleading metric. You're right that different scale needs different techniques, but comparing sizes of actually existing systems in this way it's very easy to conclude your strongest designers who successfully wrangled a large problem into a simple system are your weakest designers because what they designed is so small.