This post’s argument boils down to the classic lump of labor fallacy. Human needs are infinite and the low cost of distributing software means that over time software will meet an ever greater number of those needs.

It’s very probable that we’ll be needing to operate and maintain 10,000x (or higher, if the cost of building said software falls more rapidly) more software in 2060 than today. And no matter how good the tooling gets, we’ll never fully automate the creative aspect of engineering and problem solving.

I am looking forward to the day that more people can participate in the software creation process and reap some of the amazing economic benefits though.