I think that now the bottleneck will be less the tasks after development but before. Deciding what to build will be the new bottleneck. Finding ways to measure the system to make decisions, all of these need to improve. We risk building the wrong thing at an accelerated pace