[colechristensen]

I think you are quite wrong, at least in the narrow sense covered by SAE competition engineers (and areospace engineers).

Recruiters look at grades and activities. If all you have to show for yourself are perfect grades, it's entirely possible (and sometimes way too likely) that the only thing you're good at is getting good grades. Add in experience that you can prove yourself by talking about and suddenly grades matter a lot less – you still have to pass your classes and understand the material, but the experiences prove your competence in combination with the grade.

Not everyone developing a product has to be an engineer, so if you want a lot more technicians, fine. But you still need a good portion of engineers with the full depth of knowledge. People who program don't necessarily understand this, because a lot of programming work is technician work that you can teach yourself. Not that it's impossible, but very few people self-teach partial differential equations, and it's not something you can just pick up via osmosis.

[eeZah7Ux]

Perhaps software engineering needs more engineering and less "stack-overflow technicians".

Yet, the need to gain extensive knowledge does not require the traditional university-then-work model.

[colechristensen]

I could see 3-year degrees, but that's it.

Having gone through an engineering program, the first two years were prerequisites which were entirely required before anything useful was learned starting in year three.

And it wasn't just extensive knowledge gained, that was the least important part. It was learning how to think. There was quite a lot of mindset change. It just isn't feasible for many kinds of engineering to babysit high schoolers for 5 years before they become useful. (undergrad graduates aren't useful when they start either).

[gaius]

It was learning how to think. There was quite a lot of mindset change.

Indeed. I remember when a prof introduced us to linear programming - it was literally mind expanding - suddenly a whole class of problems that I couldn't even properly think about before because I lacked both the vocabulary and the mental structures, were within my grasp. Now it would theoretically have been possible to teach myself LP... But I couldn't have in practice, because I didn't even know there was such a thing, let alone what it was called. All the googling and stack-overflowing in the world can't help you in that situation.

A sufficiently motivated student could do it via MOOC sure (I am doing one myself right now to get up to speed on ML) but for most people, most of the time, the traditional old-fashioned in-a-lecture-theatre degree is the right call.

[eksemplar]

I don't think there is anything inherently wrong with google programmers. Sometimes looking something up is faster, sometimes it's inspiring.

The thing a degree does though, is make sure people know what stackoverflow code to ignore.

You can certainly obtain the same knowledge on your own, and you can certainly avoid getting it with your degree.

The thing is though, hiring people is about minimizing risks. The most expensive mistake you can make as a manager is hiring the wrong person, and degrees offer the highest probability of skilled labour.

[jlarocco]

> I don't think there is anything inherently wrong with google programmers. Sometimes looking something up is faster, sometimes it's inspiring.

I think the term "google programmer" covers too much ground to make a blanket statement.

On one hand, Google and StackOverflow can serve as an easy to use interface to documentation and troubleshooting information. I think this use of Google is great.

On the other hand, it's also possible to use them as a source of copy/paste code snippets without necessarily understanding what the code is doing. This second kind of "google programmer" is a huge problem. These are the type of people we all laugh at for not being able to solve FizzBuzz (assuming they haven't found somebody else's solution to copy/paste).

> The thing a degree does though, is make sure people know what stackoverflow code to ignore.

I wouldn't count on it. Even if a code snippet on SO is 100% correct, that doesn't mean it'll be the best solution outside of the original asker's situation.

[eeZah7Ux]

Please understand my sentence in context. I did not write that a search engine should not be used per se.