President Obama calls on every American to learn code

> Unless it's part of your job, the odds are vanishingly small that it will ever be of much use to you.

There are very few, if any, areas of knowledge work (which is broader than STEM, but subsumes it) to which a small amount of programming skill cannot be productively applied -- even the ones where institutional IT policies have restricted the available means of applying that skill to, say, Excel macros.

> Nor is it something that's useful in its own right because it promotes clear thinking. Nor is it a steppingstone to other, more broadly useful skills.

I'd have to disagree with this pretty strongly. There are probably ways you could develop some coding skill without learning broadly-applicable problem decomposition and analysis skills that are usable almost everywhere and not usually taught in other contexts, but you'd have to make a special effort.

"encouraging curiosity" this.

I've tried to talk to younger kids in my family about building things with computers. You'd be surprised how hard it is to get them interested, even if they play Minecraft all day.

For many of us, we got interested in programming though randomly hacking a game at one point in time. Telling kids that they can build their own games or apps and helping them potentially get interested is a big deal in itself.

Yes it is, depending on where you go to school. 50 states are not all the same.

Large employers (or any other corporations) don't see labour costs as an end in it of itself, but merely means to an end: the ends being (whether in the long term or in the short term) making money. They could pursue the ultimate "throw more bodies at the problem" strategy, but per Brooks it is known not to work; companies that do so are penalized by the market.

On the other hand engineers can't choose between working many low-satisfaction jobs simultaneous or working for one high-satisfaction job -- so they choose jobs that maximize their satisfaction. While there's low marginal value in higher salaries past a certain point (so other factors, e.g., for me it's working on interesting things take over); however, significant sub-market compensation (in terms of both salary and stock) -- compared to what an individual of similar ability, experience, and location would get in a company of similar size, age, etc... -- does greatly decrease satisfaction.

That gives: engineers want to maximize their satisfaction which requires not being underpaid; companies want to maximize the money they make, but can take many different approaches to this.

So at the end, the market "chooses" to hire fewer, but strong engineers through a combination of university recruiting (and paying high salaries even to new graduates), tolerating a certain degree of churn among early-career software engineers, and also aggressively going after a smaller cadre of experienced programmers who can also mentor talented but less experienced new-hires and interns.

So no, I don't see the scenario you describe as happening. Instead, I'm not alone (http://en.wikipedia.org/wiki/Average_is_Over) in thinking that similar trends will occur (or, rather, are already occurring unseen) in other white collar jobs, as well as many skilled blue collar jobs. Company finance departments will want to hire workers who are able to write VBA and not rely on clunk recorded macros; manufacturers will (and already do) want to hire workers who can program a CNC machine[1]. This will not make them software engineer any more than me being able to solve a differential equation makes me a mathematician or understanding logic circuits makes me an electrical engineer (to take it one layer lower, knowing Maxwell's field equation and even using it on the job, doesn't make an electrical engineer a physicist). Circling back, this is exactly why it makes sense for people to learning programming -- if robots are automating large parts of people's job's, there will jobs that require knowledge of both the job and programming, to be able to write "userland" programs on the robot, as oppose to program the robots' embedded systems (or to write Excel VBA scripts as opposed to build Excel itself, etc...)

[1] Language they use is G-code (http://en.wikipedia.org/wiki/G-code -- note the date, it's nothing new), and yes, it's Turing-complete: http://callenblogi.blogspot.com/2011/03/bf-interpreter-in-g-...)

When I think about the practical skills that I learnt in my education, years later there were only a few key concepts which really mattered: rudimentary arithmetic in Mathematics, good writing/reading skills from English, understanding of empiricism from the Sciences.

I think the question is: is there a foundational skill or concept that learning coding gives you?

I think perhaps there is and it's something to do with the application of symbolic logic, however I am not sure whether I had this before or after I started coding. (I do, however, feel that I grasped it far better from constant practical application than I did from mathematics lessons.)