Among other factors mentioned: employer-specific skillsets.
Technology jobs are very highly specialised. And I'm not just speaking of computer tech.
I'm old enough to have seen several generations of obsoleted "gold tickets". Nuclear Engineering, future destroyed March 28, 1979. Freeway engineers, the freeway protests of the 1960s/1970s. Big Three auto manufacturers, dream destroyed with the first oil embargo, and sudden onslaught of more fuel efficient, less expensive, and better quality imports from Japan and Germany. Petroleum engineers, dream destroyed as the oil market collapsed in the 1980s. Aerospace engineers, dream destroyed with the fall of the Soviet Union and redundancies in the military-industrial complex throughout NATO powers.
Travel agents, newspaper journalists, and Perl hackers are only more recent victims of the same levels of specialisation.
It's not that these people are dumb, many were brilliant. But they're also laden with a tremendous amount of arcane knowledge, that, unfortunately, gets in the way of learning new skills.
It's been painfully obvious for a long time that, individual success stories notwithstanding, retraining and "reskilling" typically does very little to alleviate this.
Yes, future generations get trained into new skills, some of which may still be germaine 4-8 years after they've completed schooling. But my observation is that job and skill cycles are iterating faster.
So that's a part of it.
Yes, there are other dimensions. Obligations, family ties, social ties, fears, the fact that the most mobile do jump ship first, and more.
But really, it's quite crushing. You may not understand it now, but you almost certainly will, eventually.
Regarding skill sets. I have only been in software 12 years or so, but I noticed a type of tech 'gamble'. Basically, when you decided to learn something in tech you are basically gambling that the tech will last. In 2007 many faced a simple decision: AS3 or iOS. Supposed you chose AS3. Then, worked on it from 2009-2012, quite profitably. Currently the AS3 job outlook is horrific, and all that knowledge is a waste. Now, iOS people are skilled, wanted and are taking lead roles. (Oh yeah, now you are old). Want to learn GO? Be careful.
There are gambles, yes, but they are microscopic compared to retraining from one kind of engineer to another.
If you have n years of AS3 experience (n=7? Or what happened between 2012 and now?) and find yourself unable to compete with people with the same n years of iOS experience, I humbly submit that something else is wrong. Never mind that AS3 is a dialect of ECMAScript, which I hear is still used in a few remote pockets of the web, that hard things about programming isn't about the language and the framework, it's managing complexity. I struggle to see how retraining to iOS is going to take more than a few weeks of systematic reading and practising (possibly taken on your own time), and little more than a few months on the job to be on par with people who have several years of experience.
But more important is all the stuff that isn't about programming at all: Being good at organising your work, communicating well, testing, making sure you understand the entirety requirements as they are intended and that your delivery fully satisfies the requirement etc etc. Someone with n years of programming experience should master those things, and these count very heavily when picking people for lead roles.
Anyway, bottom line: the difficulty of switching from AS3 to iOS doesn't map to the difficulty of switching from nuclear engineering to environmental engineering.
Quite. I like the "Tech Gamble" concept. May well borrow that.
This also reminds me of a back-rationalisation I've seen for Linux adopting the penguin as mascot -- not the Australian zoo killer penguin story.
Antarctic penguins, when deciding where and when to enter the water mill about on the edge of the ice. The problem is that the first one in is taking a significant risk -- leopard seals, which eat penguins, may be lurking.
But once the brave pioneer takes the plunge (or rather more likely, is pushed, jostled, or slips in), the rest of the flock follows. There's safety in numbers ... and fish to eat.
I chose the things I learn not by what I think might be most profitable, but what I feel is most fun. If I wanted a job with boring but profitable technology I'd learn COBOL or MUMPS.
I agree that switching from nuclear scientist to something else might be hard but with programming languages after a certain amount of years the gamble is gone.
The core concepts of every language are the same.
If you acquired a certain amount of experience, switching is just looking for best practices in your new language. Of course it will take some time to ramp up but within a month you are somewhat productive within 3 months no one will see the difference.
In up markets it's easy to say, "I'm a mobile developer" but in a down market, only iOS will do. You're right there are gambles.
I see 4 ways to mitigate this:
1 - Live somewhere that has many companies or be prepared to move.
2 - Keep in contact periodically with everyone who respects your work. (Don't make the "Hey it's been 5 years" call one where you're asking for a favor)
3 - Always keep learning.
4 - Communicate in your customer's language. If you are strictly an ActionScript expert, you are very narrowly pegged if you haven't learned iOS on the side. But if the buyers of technology view you as "A mobile expert who makes brands come alive across many devices" then you can survive technology switches.
4.5 - Help others before you need help.
In the grand scheme it's still better to be a laid off technology worker than a laid off steelworker. I still feel their pain though - it sucks to lose your job, especially if family is involved.
There's a professor, Economics at Harvard IIRC, who points out some telling characteristics of cities in this regard. Those with large numbers of smaller firms are far more creative in the long run than those with small numbers of large ones. Specific contrast were NYC and its garment district vs. the large industrial cities of Pittsburgh (steel) and Detroit (autos). Ironically, it's the cities which most successfully employ division of labour and assembly-line techniques which are penalised most harshly.
Link in a sec....
Benjamin Chintz came up with the theory. Edward Glaeser is the lecturer.
Perhaps IBM has more workers that are like the workers at AT&T than we would have expected. AT&T is in a difficult position as its extremely large workforce is being attacked into irrelevance by new technology upstarts.
> To Mr. Stephenson, it should be an easy choice for most workers: Learn new skills or find your career choices are very limited.
“There is a need to retool yourself, and you should not expect to stop,” he said in a recent interview at AT&T’s Dallas headquarters. People who do not spend five to 10 hours a week in online learning, he added, “will obsolete themselves with the technology.”...
By 2020, Mr. Stephenson hopes AT&T will be well into its transformation into a computing company that manages all sorts of digital things: phones, satellite television and huge volumes of data, all sorted through software managed in the cloud.
That can’t happen unless at least some of his work force is retrained to deal with the technology. It’s not a young group:
The average tenure at AT&T is 12 years, or 22 years if you don’t count the people working in call centers. And many employees don’t have experience writing open-source software or casually analyzing terabytes of customer data."
That's a great top-down message from the AT&T CEO.
But the bottom-up reality is that people coming into work at AT&T pushing open source, cloud, or data science agendas are perceived as threats, and are managed out.
It is a really nice thing that they are taking the right steps, but personally, I am very very pessimistic. Just encouraging people to take classes is one thing, but them actually applying those skills to solve the problems that at&t faces...that is a different question.
I really really do hope that it works out though; my rational mind believes it otherwise.
"Technology jobs are very highly specialised. And I'm not just speaking of computer tech."
I would argue that computer tech is kind of an exception.
I can see a Nuclear or Aerospace Engineer having trouble adapting his knowledge and skills to other professions, since that knowledge is deeply tied to the domain, but a Perl (or C++/Java or whatever) hacker should't really have too many problems in transitioning to another language.
That they would find difficulties in finding a new job because dumb HR recruiters see 10 years of Perl experience in the resume as opposed to the "10 years of Swift" (sic) they are looking for is another matter...
Yes, there's a fair bit of specific domain knowledge, but it's also built on top of a bunch of common foundations: physics, calculus, chemistry, differential equations.
And the tools in much of technical work have been converging for years -- technical and scientific computing require rudiments of programming (Fortran, still, or more modern languages), source code management, project management, and a heavy helping of sociological people understanding (look up Charles Perrow's books). That can transfer, at least in theory.
Whether or not it does, of course, and whether or not firms, organisations, or government sector are interested in hiring is a whole 'nother question.
I'm reminded of a video on the old Soviet rocket motors with boosted performance (the fuel pump turbine exhaust was vented into the rocket combustion chamber rather than offboard, for another 10% or so gain in thrust) -- though attempted use resulted in an explosion only a few seconds after takeoff and loss of the mission.
In the video, however, several former Soviet rocket scientists talk about their experience after the Soviet space program was shut down. Rocket scientists, literally, with no place to turn for employment. That's where you start thinking about inefficiencies in usefully allocating talent. Neither the Soviet nor American economic systems seem to have a particularly good answer for that. Though in the US, Wall Street seems to have a capacity for sponging up some of the potential. I don't find that particularly useful.
The Egyptians built pyramids. I wonder if that wasn't in part a large skills-retention and problem-solving exercise.
I agree. Actually it's interesting to see
how IBM is trying very hard to stay relevant as far as programming languages in their products. If you look at the video in this "API Connect" demo, you can see support for microservices in Ruby, Python, PHP, Java, Node.js, Go, and Swift. As an old time C/C++ guy it's difficult to determine which of these are fads. Also no C# support, but maybe this will change with the opening of .net.
>It's not that these people are dumb, many were brilliant. But they're also laden with a tremendous amount of arcane knowledge, that, unfortunately, gets in the way of learning new skills.
I don't use the term "brilliant" for anyone who can't learn outside their narrow specialty.
There's a middle ground to what he's saying. Once you MASTER something that all your life has been in demand and incredibly lucrative, it becomes hard to believe that completely pivoting away from that knowledge is your best course of action. I believe in most cases these down turns seem temporarily, a short storm to weather with a small reduction in pay until things get going again (see Big 3 auto). Unfortunately things just keep spiraling downwards and by the time you realize it, it is too late.
Definitely such short sightedness and stubbornness can be associated with brilliance.
It's not a matter of not being able to learn. Of course they can learn. It's a matter of not being able to instantly re-specialize in the next "in demand" skill.
1. The opportunity costs for switching away are high. I suspect this comes into play in many areas -- say, emigrants to a new country who never develop a high functioning level of language skill (something I'll add applies just as much to English speakers abroad as those moving to English-speaking countries, if not more). And yes, there absolutely are exceptions, often many.
You're highly effective in using one language. You operate functionally at the level of a young child in another. Cross training to the new language is a massive investment of effort. Adult minds lose plasticity present in children, so the task is inherently harder. Plus you've got to do everything else necessary to keep going. It's a Red Queen's race.
Shifting tracks requires a very real sacrifice both in increased effort and reduced reward. It can be short-term satisficing to continue stretching out the lucrative current skill so long as possible. Situational awareness of that dead end may not dawn until late in the game as well.
2. Some skills are synergetic -- knowing one thing makes learning another easier. And yes, I see this all the time, particularly in metaphorical spaces. Thinking is I strongly suspect very much a matter of analogs, and collecting and applying appropriate analogs helps.
But ingrained knowledge is ... quite different. Changing handedness, or quitting habitual behavior (drugs, smoking, alcohol, gambling) is exceptionally difficult.
YouTube's "Smarter Every Day" has a great segement on learning to ride a backwards-engineered bicycle headset. Using gearing, the response of handlebars and wheels is reversed. It's all but impossible to ride this bike if you know how to ride a regular one, though you can learn, with sufficient time.
But: learning to ride the backwards bike makes riding a normal bike impossible. Until you've learned to context-switch between the two.
I've been simultaneously waiting for and dreading the same future for rails engineers, since so many are under the belief that web development is the ticket...
Yeah, I've started seeing occasional queries on HN about what the Next Big Thing will be. Which raises another challenge: chasing the tech curve. It's one thing to have found the right tech, another to correctly anticipate the market's next move.
The worse when you realise everything that's up-and-coming is recapitulating all the mistakes you recognised from the last time.
This is why I think learning new marketable skills should be everyone's hobby. Also, in our field or any professional field, you really should be very financially secure after 50 so that a job loss doesn't hurt you.
The problem is learning new skills is kind of a catch-22. After you've been doing something for a few years, assuming you're reasonably intelligent, you get to be an expert and command expert wages. Your employer is paying for an expert. He doesn't want to pay you expert wages to learn a a new, marketable skill. He can hire someone who already knows it. Or he can hire a new college grad for a fraction of what you make.
So you learn on your own time. But since you're doing tutorials and what not, and your time isn't infinite, when you actually apply for a job with this new skill your potential employer is going to wonder why he should pay you expert wages for a skill with which you still have effectively no professional experience.
You can, but only if someone's willing to hire you. Employers are leery of hiring you for substantially less money than you made in the last job. The fear is you're still looking, and a few months down the road (before you've really accomplished anything) you'll quit.
I think a much more likely scenario is that you won't get an interview at all. An employer is unlikely to consider you for a junior role that requires < 5 years of experience when you've had > 10 years experience in another field.
Software is highly unusual in that you can learn marketable skills as a hobby. It's not like a chemical engineer can learn marketable skills in civil engineering on the side. And not everyone can do software for a living.
There is that advantage, though some skills are harder to pick up (stuff relating to hardware, commercial software packages, etc.). This also means more competition as there are fewer barriers to entry.
That's a nice dream but most people are probably too busy staying relevant in their current profession to be concentrating on learning skills for another career that may never happen.
Exactly. Or there's no time. I get up at 6:30AM to work and get home around 8:30PM exhausted. Maybe get an hour or two to play with the kid, then it's off to bed for all of us. When I do actually have a weekend, it's spent taking care of all the shit that didn't get done during the week. Exactly when do I sit down to learn AngularJS?
Except that the rate of change is picking up so the likelihood that the value of new skills is diminishing as quickly as you learn them. We need to refocus society on providing a minimal viable income.
Up-and-coming tech frequently has a flood of talent. Dying: companies with vested systems (say, Cobol, or Solaris) are catching the tail-end of the curve. The truly good programmers have gotten out, or switched skills, or transferred to management.
Hopping on a sinking ship usually doesn't last long.
(Nothing against Perl. It's just a seriously fading language at this point.)
This is a very interesting angle of discussion, really
But I think this is partially their fault as well
A Nuclear plant engineer, while specialized in some element of that plant (and there are many elements, in civil, mechanical, chemical engineering) has knowledge that's applicable in a lot of different situations
Were the Automobile engineers content of designing more of the same poorly constructed vehicles? While the ICE is old technology I can't imagine the things that could come out of it if they really invested in rethinking it (also other parts like the drivetrain)
> But they're also laden with a tremendous amount of arcane knowledge, that, unfortunately, gets in the way of learning new skills.
That's true. But I think it's also a matter of attitude
You know, I can be the best specialist in some technology, let's say: installing Oracle, or even better, let's say I'm a top notch Golang programmer today
Tomorrow Golang won't exist, or won't be relevant. I am SURE of this
> But really, it's quite crushing. You may not understand it now, but you almost certainly will, eventually.
It's really, really easy to think you're one of God's Chosen (or Allahs, or Jehovas, or Buddha's, or ...) when you're at the top of the world, you've got a multi-trillion-dollar wind in your sails, you went to the best schools, and you make titanium and airaluminium dance on the wind and all that.
1980s aerospace engineers weren't the first to take that tumble. They won't be the last.
> It's really, really easy to think you're one of God's Chosen (or Allahs, or Jehovas, or Buddha's, or ...) when you're at the top of the world, you've got a multi-trillion-dollar wind in your sails, you went to the best schools, and you make titanium and airaluminium dance on the wind and all that.
Fortunately nobody posting on Hacker News threads like these are relatively new engineers who think they're God's gift!
Technology jobs are very highly specialised. And I'm not just speaking of computer tech.
I'm old enough to have seen several generations of obsoleted "gold tickets". Nuclear Engineering, future destroyed March 28, 1979. Freeway engineers, the freeway protests of the 1960s/1970s. Big Three auto manufacturers, dream destroyed with the first oil embargo, and sudden onslaught of more fuel efficient, less expensive, and better quality imports from Japan and Germany. Petroleum engineers, dream destroyed as the oil market collapsed in the 1980s. Aerospace engineers, dream destroyed with the fall of the Soviet Union and redundancies in the military-industrial complex throughout NATO powers.
Travel agents, newspaper journalists, and Perl hackers are only more recent victims of the same levels of specialisation.
It's not that these people are dumb, many were brilliant. But they're also laden with a tremendous amount of arcane knowledge, that, unfortunately, gets in the way of learning new skills.
It's been painfully obvious for a long time that, individual success stories notwithstanding, retraining and "reskilling" typically does very little to alleviate this.
Yes, future generations get trained into new skills, some of which may still be germaine 4-8 years after they've completed schooling. But my observation is that job and skill cycles are iterating faster.
So that's a part of it.
Yes, there are other dimensions. Obligations, family ties, social ties, fears, the fact that the most mobile do jump ship first, and more.
But really, it's quite crushing. You may not understand it now, but you almost certainly will, eventually.