Well, other than that he's completely mistaken about the career advice he offers here. I'd be very curious to know if he's ever been the highly-respected programmer in a science lab, or personally knows anyone who is. My science friends tell me the job doesn't exist; any code that needs writing, a grad student will do it sufficiently well (on top of their lab work) for no more money.
I’m not brave enough to speak in extremes/absolutes like "completely wrong" on a topic like this.
My experience aligns with his advice. As a finance student, I was selected for a research project because of my database/SQL knowledge, which got me mention in the Journal of Finance when my prof. was published, and which later paid unexpected dividends. Later, knowing to code in a finance/analyst job ultimately positioned me for a startup that has left me much better off financially than if I had stayed in my job. I also have a friend in the pharma research field that did very well because he could write code.
(Note that it’s not always the money that is the pay off. Sometimes it is an opportunity that leads to other opportunities that would have never been open to you had you not been able to code, as in my case. )
So, in my experience his advice has proven to be right on.
I worked in academia for a decade and the code written by Hgrad students is a problem... but 95% of faculty are so clueless about computers they have no idea. And decent grad students know it.
I've worked in physics and chemistry and I've seen this code... It is like a ten year old with cheez whiz sealing your bathroom. It works for a while, but it is not professional code. Smart domain experts know they are just barely passing with respect to coding solutions in their field,
A kinder interpretation is that he meant: be an expert in that field who can also code.
Which is probably pretty true already, much as everyone can use a word processor or spreadsheet. Programming to get a job done is not that hard, especially with languages like python. Over time, the tools will get even easier.
Being a tool-maker seems to have some mileage left - but that will also be made easier over time.
EDIT it's hard to predict what will happen, but for several decades, programmers earned above average wages. Cool... but weird. I saw a great explanation, that during that time, not everyone who wanted a computer had one, and each transition (mainframe/mini/workstation/PC) increased the market size, and therefore the demand for software. Higher demand --> higher wages. But according to this explanation, PCs maxed out their market around the year 2000. While this mightn't be strictly true, as the third world is unconquered, and people buy second and third devices (e.g. smart phones), the rate of increase has slowed.
Those above average wages will never return. There will just be average wages. But if you like programming, that's OK.
What I believe he means is that since code is the tool, to code within the programming industry is laboring on the tools themselves, but to code to solve the specific problems of another domain is applying the tools.
Ability to code in another domain gives you leverage for work done IN that domain. The caveat is that knowing ONLY how to code without working knowledge of the domain itself will not get you very far.
"People who can code in biology, medicine, government, sociology, physics, history, and mathematics are respected and can do amazing things to advance those disciplines."
Are you talking about anything in particular when you mention that? An example would really help your case here.
This jibes with my experience. Once upon a time I know a lot of grad students in the sciences, once they found out I was a professional programmer the fun was effectively over due to a barrage of programming questions. They all hacked on some sort of code.
