Basically I told my then boss see you after Christmas - If I have followed that advice I would have stuck with Oracle and Java or spent my career in Mainframes
And last I heard, there are tons of job openings asking for experts in Oracle and Java.
So how exactly would that have been terrible in terms of getting a steady income?
Trouble is if your company pivots and you have been doing Mainframe COBOL for 20 years - you might find transitioning to a new language hard.
Now? I'm stuck working for a salary in a job that pays the bills. I live in Texas (I know, I know) where IT salaries are already low. Texas has that "right to work" mentality. My bosses are not real IT guys in the sense that they love IT. They're in it for the salary, whereas I'm in it because I still love it after a span of pushing bits over three decades. I would retire only if I won the lottery, but if I did that, I would likely buy a Z-Series mainframe and spend my days playing.
Then you are not following the advice discussed.
At least, that's my interpretation of the article's logic mixed with what appears mainstream to me. There's a lot of room for subjectivity, but it's difficult to argue that Java/Android isn't an extremely mainstream direction.
It's actually a pretty good point and is a decent way to always remain relevant. You can't be left behind if you're always on top of the latest thought trends.
EDIT: A missing piece the author points out too: don't be an early adopter of a thought trend (Point 1: Forget the hype). Only jump in when it becomes mainstream. If you adopt something before it becomes mainstream, there's a chance it can fail. If you wait until it is mainstream, but get in just as it becomes mainstream, you get the benefit of being an early adopter and the benefit of never being on the unpopular viewpoint.