>Part of me hopes software development is the next job we automate away

We are hard at work automating software development. Thanks to massive advances in programming language design, IDEs, the invention of module systems in combination with the massive success of OSS in that space and the proliferation of massive frameworks, something that took a year of work two decades ago is now at best a weekend project of a much less skilled person.

The thing that keeps us all employed is the incredibly elastic demand. And of course that the work to find and describe requirements is unlikely to be automated away anytime soon (and already often takes more time than actual implementation)

> Thanks to massive advances in programming language design

What massive advances in programming language design have we had since 1998!?

While a lot of the real advances in language design are not new, popular adoption of stuff like functional programming has happened in large part in the last twenty years. E.g. list comprehensions in Python were not introduced until Python 2.0 (2000, which just squeezes into your range). And Java is starting to get similar features, which is great for companies who have invested in the language.

> Some people will still pick wrong for a multitude of reasons. Hopefully, those people can pivot.

We are lucky as software engineers, but I was also making the point that if your line of work gets automated, hopefully you can pivot. This applies to everyone, and is similar to when you are no longer enjoying your field of work. Changing careers is not always an option, but let's not forget that it usually is when we talk about automating an entire industry and putting them out of work.

So give them the support they need to retrain. We're creating so much value with these automated systems - who is the beneficiary? Surely the compassionate thing would be to use some of this resource to help those that have been trodden on by progress/

I'm hoping for a future where the machines do all the work required to provide everything human beings need to survive, AND where we've figured out a reasonable way to distribute resources in a world where not everyone will be able to "contribute to society" by spending huge amounts of their waking time doing something they'd not otherwise do, in exchange for tokens that can be redeemed for food and shelter.

Yes, but that would be communism, so let's just let people who were born into or stumbled into their wealth amass more of it. I think in general we need to start being more thoughtful about how the things we do can sometimes be hurtful to billionaires.

Well, while you’re joking the original idea of communism and what I’ve been taught about Lenin was about excatly this - communism is inevitable because of technical progress. Call it what you want - universal income, handouts, we will end up in communism.

And yet all implementations of communism has failed so does this mean technical progress eventually leads to societal failure? Also I believe we will end up with socialism, not communism.

> And yet all implementations of communism has failed so does this mean technical progress eventually leads to societal failure?

Only if you believe that once an attempt at something has failed, all future attempts will also fail.

> Also I believe we will end up with socialism, not communism.

Socialism is worker control of the means of production, and is an intermediate step along the route to communism, a fully classless and stateless society.

So if your line of work gets automated, you retroactively "picked wrong"?

It's not far from it. Software development is in small company with respect to the need for constant re-education and skill maintenance. The vast majority of developers with skills from just 10 years ago are wholly unemployable right now.

You mean somebody who knows how to develop in Java, PHP, Python, Perl, C, C++, Javascript, Ruby, LISP - all languages developed more than 10 years ago - is "wholly unemployable right now"? Doesn't seem to match my experience even a little bit. Programming practices also hasn't changed that much. In fact, there are still jobs for COBOL and Fortran programmers - not that many, but there aren't that many COBOL and Fortran programmers either nowdays.

There's a shortage of COBOL developers that are needed to maintain and update legacy financial mainframes that still don't make economic sense to replace since they operate at critical infrastructure points 24/7 and over the years have become pretty bulletproof.

Replacing these systems is extremely difficult since they're so relied on a replacement needs to be nearly flawless form the start. If you stumble upon on a bug in your international wire transfer code you could be messing up a $10 billion international aid wire that was meant for hundreds of factories.

Um, no.