The mindset of “infra can’t support billions so let’s build that first so we can thn build apps that will attract billions” is the wrong flow of events 

A better flow would be: apps are built first, strains existing infra, better infra gets built, apps scale alongside it 

So the infra is build out of the very real demand. Compared to the infra being built because of the assumption of future demand.

Could AWS have been built without Amazon itself needing more capacity, realising they could do it at scale much cheaper thn solutions at that time? Yes. 

But would it have made sense for Amazon to build that before Amazon itself and other firms reached that scale? No imo.

You also get more relevant infra being built when it’s done in that sequence. Builders get more an accurate understanding of what parameters to optimise for and what kinds of infra to build because you have a real understanding of what end users need.

marketing, product and growth | prev content & comms lead at @nansen-ai

Build apps first, strain existing infra, build better infra.