[Analemma_]

And those who say that would be full of shit. We have doctors, lawyers, judges, scientists, etc. all remaining productive and employable late in life, and those fields also see change that you have to stay on top of. The ageism in programming is a desire to avoid workers who have experience to see through the “work 100-hour weeks and you’ll get rich!” propaganda, and this explanation is nothing but a fig leaf.

[JamesBarney]

I don't think those jobs change nearly as much as software. Winforms development on VSS/windows and react/node.js work on github/AWS have far less in common than what an allergist was doing 10 years ago vs today.

[Xixi]

There are new medicine/procedures/tests coming out all the time, we (as software engineers) are just not aware of it.

On the other hand is winforms development that different from react/node.js? On the surface it is very different, but good UI principles -which I argue is what truly matters- didn't change all that much, if at all.

[copperx]

I would venture to say that 90% of what doctors learn in med school does not get outdated. That 10% that changes, at least for non-research doctors, is just about new treatments, new drugs, new protocols and algorithms for certain diagnoses, especially the ones that weren't treatable to begin with. A doctor in her 20's will treat 80% of the diseases in the same way when she's 70. Cognitive decline will not make a doctor less effective. Crystallized intelligence is valued.

The theory that you learn as you work towards your CS degree does not get outdated either, but most of what you use in the average software engineering job is not about theory. It requires knowing the latest languages, frameworks, and practices. These things change every six months. Even if you are not moving to a new language, the language is evolving. Even if you keep using the same framework, the API will change. Crystallized intelligence here is mostly useless.

[robbiep]

> I would venture to say that 90% of what doctors learn in med school does not get outdated.

> A doctor in her 20's will treat 80% of the diseases in the same way when she's 70.

this is patently untrue. I'm in my 6th year of practicing Medicine. During my first day of classes, we were told that 50% of the knowledge we gained would be obsolete by the time we graduated. I took this as a metaphor for the pace of change. However, with the benefit of 2 cycles (Medical School, and then my working years) this has been absolutely true.

We have totally changed both how we approach heart disease/heart attacks (classifications) and treatments. We have a cure for hepatitis C. Dietary recommendations have been upended. Treatment of trauma has changed. That's just a brief overview.

Whilst the underlying biology may remain the same, treatments even over the last 6 years have changed significantly, and continues to change significantly

[JamesBarney]

So googling most common diseases, the top results were diabetes, heart disease, cancer, asthma, arthritis.

With the exception of cancer has the treatment of these diseases changed that much since 2009? If you're a specialist I imagined it's changed more but for a general practitioner it seems like treatment hasn't changed a whole lot. Most of the people I know with these conditions are taking the same drugs in 2019 as they were in 2009.

[arkades]

Yes, it’s changed. The changes are just more subtle than non-physicians have insight into. They’re technical details regarding -when- to treat what and for how long, which medications are first line now, etc.

I know that seems less in your face than what your example was, but... to a non-programmer, your job is identical to what it was a decade ago, too. You’re just typing a slightly different pattern words into a compiler than you did before. Turns out, people outside of a technical specialty have a pretty shitty understanding of the internal workings of that specialty.

[arkades]

I was at an Alzheimer’s talk the other day given by a neuro to the psych, Geri, and palliative departments.

You could tell the psychs in the audience hadn’t read up on Alzheimer’s in the last couple of years, because their questions were just so damned ignorant. And when I say “ignorant” I mean “knew 99% of the relevant information, but didn’t know about the current state of knowledge regarding diagnostic testing, biomarkers, etc.”

They knew more than they knew in med school. They didn’t know enough to competently manage an Alzheimer’s patient to modern standards. They wouldn’t know how to best (accurately, efficiently) diagnose it. They wouldn’t know when to start meds. They wouldn’t know when was the best time to discontinue meds. Most of the psych in attendance were residents - they were maybe two or three years out of school.

And Alzheimer’s has barely changed at all. That’s one of the slowest changing topics in medicine.

(Not a criticism of psych. Alzheimer’s in my institution, and generally in the medical field, is owned by neuro. They just happen to be a good stand-in for “folks a couple of years out of med school” here.)

[Izkata]

> The theory that you learn as you work towards your CS degree does not get outdated either, but most of what you use in the average software engineering job is not about theory.

My brother just started learning to program last year, and some of his questions have highlighted just how much I forgot that I had to learn at some point. So much so that I think your 90%/10% ratio does actually still apply in day-to-day software development.

[JamesBarney]

The number of new allergy medicines in the last decade was 10.

https://www.centerwatch.com/drug-information/fda-approved-dr...

[flukus]

I can see you point but it depends. Assuming a developer was half decent in the first place then moving from Winforms to React is easy, depending on what they've done along the way react is actually pretty similar to old school asp/php and adapting from desktop to web technologies is the hard part. On the other hand a lot of people that gravitated toward Winforms development weren't half decent in the first place, but that's an issue with individuals not a technology gap. VSS to git is similarly easy, assuming they're comfortable using the command line, which may be a big if.

And that's just the UI, a real developer would be doing a lot of logic behind the UI, database design, debugging production issues, talking to stakeholders, etc, that's an enormous wealth of transferable skills.

[JamesBarney]

If you had one month to write an application on the MEAN stack and all you knew about two possible devs was one had 10 years of experience on winforms, and the other had 1 year of experience on the MEAN stack? Which do you pick?

I'd pick the MEAN developer everytime.

Sure in a year maybe the second one would be more productive. But there aren't any allergists that are a year away from productivity. Even if they hadn't learned anything in 10 years.

[digisign]

I'd let the person go that chose the mean stack, in favor of someone more experienced.

[flukus]

If you had a 1 month deadline and were still hiring I'd hope the guy with a decade of experience would see the looming tyre fire and select himself out of the running.

And you're kidding yourself if you think it takes a year to get up and running in whatever GUI stack was cool this year, the guy with 10 years of experience in winforms obviously knows a thing or two about picking stable mature technologies.

[arethuza]

If I had a deadline like that then I'd pick whichever one has a track record of delivering to tight deadlines - familiarity with a particular technology stack is going to help but isn't everything.