The thing is, software engineer salaries are going down and data scientist salaries are going up and the data scientist jobs are taken by people from the sciences, who know that kind of math, that CS courses maybe are not that heavy on. So there is a great big incentive to know "maths" (continuous maths). And if one does not know, the incentive is for one to learn.
I don't know if what you are saying is true. Where is the data?
But I do believe in the long run, you will be right. Software is saturated, and a lot of the infra has been built or commoditized. The value will be there for people who can analyze data.
Too bad I don't find analyzing data as fun or as interesting as banging out a programming problem.
If you have a lot of data, but your underlying system is simple, you don't need a lot of people to analyse it. Your demand for new people crunching that data will come as the complexity of your software grows.
Think of it this way: having 10 endpoints and 10TB of plain-text data vs 1000 endpoints and 1TB of plain-text data. The later will surely require more time to be analysed, even though the amount of data is smaller.
Software is saturated and consolidated; the infrastructure has been built and is owned by large companies. And that trend will continue.
Data on the other hand is still growing, and there is enough already that has not been analyzed. Traditional software engineering will still be important, but no longer glamorous. The future belongs to the data analysts.
