"It’s obvious the game has changed" - it's obvious you haven't been in tech all that long. ;) that was a joke answer - no offence intended :) But seriously, I've heard "the game has changed" many, many, many times over the last few decades. Its true. The game is constantly changing. :)
"what you all are focusing on to stay employed in tech". - Same answer as its always been. Be reliable, show up, get stories in the Done column. Be friendly, nice, help others when they're stuck, ask for help when you're stuck, do a mixture of work, some with new tech you wanna learn, some on old legacy stuff or things other people don't like so they're grateful you did so they didn't have to and want to hang onto you 'cos you know the codebase. Look after your boss :) People forget to do this. (not being a sycophant, just genuinely making their life easier and trying to be not too much of a PITA to manage). If a job ends up being stressful / non-fun, or you're locked into all old tech with no future, then ask for a change to the role, or move jobs internally or externally.
Ha ha thanks for the kind comment. :) Well I ought to know some things, been in the industry that long. BTW, despite the above tactics I haven't become financially rich from Tech - probably could have done if that was a priority, but , that's my other piece of advice to people, prioritise happiness, fulfilment and relationships with others especially family and friends, over money, 'cos "You can't take it when you go" , and people matter more. The wonderful luxury of working in tech is, salaries being what they are, you can take the middle path, balance life and work, while no-one's gonna starve. :)
Stop following the crowd. Where the crowd is, there are lots of people after your job, and if we're honest, probably better than you.
I've had one job my entire career (31 years and counting.) It's very niche tech. I'm good at it. There are maybe 10 other people in the world who can do what I do. And they're all overworked.
The real value and security in tech is in the edges, not the mainstream.
It sounds like you found your ikigai. Wishing you continued job stability, satisfaction, income, delight, health and peace. It's wonderful to read things like this.
For anyone else reading who is unhappy, uncertain or struggling, I hope you find the same for yourselves too. I probably don't know you, but there's complete strangers out there who want good things for you too. Don't give up.
Curious how this works out financially though. Following the crowd got people into faang with 500k incomes. Working on something obscure I haven’t seen payoff the same.
It varies. Some specialist skills pay -really- well. You don't really hear about them because they're, well, niche.
Equally, yes, some folk rose up through the ranks at fang (or better yet fintech) and make mountains of loot. For each success there's a fair pile of failures though.
To rise up through those ranks, typically you need to make sacrifices. Long working hours, minimal time off, stand-by weekends, emails at night, and (imo) suffering the bs that comes with corporate jobs.
Building your own business are all those things too, but without the corporate bs.
The interesting thing about 500k though is asking what you can do on 500k, you can't do on 300. Or 200. Or in lots of the world, 50.
Time left for relationships, children, holidays, other interests and so on are important to me. So I'm prepared to balance those with raw income. I'm not making anything like 500k per year, but I've turned down fang recruiters because all the money in the world can't make up for what else it would cost me.
You don’t really have to rise up the ranks to make 500k. That’s a competitive offer for a senior engineer or where a senior with a mediocre offer will end up with stack refreshers and stock appreciation in a couple of years withouts any promotion. It sounds wild but is true.
Also, many of the roles are just 40hr work weeks. Not that much more stress than a 150k a year job in my experience.
You are right that after a certain point more money doesn’t give you more utility in the present. However, it does buy you time in the future given the capital that you are bound to accumulate and compound.
I think what GP meant was that money, beyond a certain point, means little. Upto that point money does mean a lot, and anybody who denies that has never lived in poverty.
What that specific point is, depends on who you're asking. If you ask me (a student in India), that point lies around $50k/year. For me that is an unimaginable amount of money, and I don't believe earning any more than that would result in a significant increase in quality of life.
But if you ask someone in USA, you might get a different answer. Maybe for them $300k/year is around where earning more wouldn't necessarily result in a significant benefit.
It depends on where you are, who you've been around, and what dreams you have.
Are you that engineer that fixed that AS400? Back in the early 2010s, there were like 5 people across, the whole UK who could work on those mainframes.
Not me :) - but that's the sort of thing I have in mind. I remember mainframe folks in the run up to y2k getting literal "blank cheques" to work on stuff.
I would be really interested to hear about what you do. If you don't want to share it in public, you can email me at the address in my profile. I'm always looking for a niche, but I haven't found one that has stuck.
Competition never hurts :). Quite the opposite, I spend a lot of time training others in the field, and passing work in their direction.
I can't go into detail here though, as that would dox me, and I'd prefer my HN person not to mix with my IRL person.
The thing about niches is that you don't really "look for them" - you stumble over them as your skills develop. You -could- learn my job, of course, but it would take a lot of time and effort.
Part of what I'm saying I guess is that I am where i am because of a long career, amd I've accumulated a lot of experience, which is hard to pass on.
i know what you mean with not going into detail. i only need to list all my favorite programming languages for someone to be able to identify me.
but, if i may ask, is that work that can be done 100% remote? would it make sense for an older software developer to pick that up? i am a generalist with lots of experience and i'd love to train up in a niche field to help me find work.
if you think that makes sense id appreciate if you could ask your IRL person to email me :-)
Depends on uour definition of "older". In this niche, no, being an older worker would be detrimental. The root underlying problem (in any niche really) are people aging out the system, leaving a lot more work to be done by fewer people. 40 would be OK. 60 would be too old.
Clearly you -could- learn it now, but the effort would be immense. In the same way that you could learn to be a doctor, but at some point it becomes too much like hard work.
Also, unfortunately, there's a cost to entry. It's proprietary software, and you'd need around $5000 to get kitted out. Plus a lot of time.
> if you think that makes sense id appreciate if you could ask your IRL person to email me :-)
He just explained that he has no intention to dox himself plus with 31 years of experience he is one of the few in his field. Do you think you would be able to catch up and be able to stand your ground in this market with less than say 10 years? It wouldn't make sense and you probably wouldn't risk doing the change so even if he thought it made sense to send you an email, it would be for nothing. You can still get his advice and use it for the specialization you can get based on your current experience.
sending a private email is not necessarily doxing himself, at least not publicly (which is what doxing usually implies)
besides, the email would not have to even reference this thread. anyone reading who is in a similar situation could be responding and i would not be able say if it is the same person.
for the experience part he said he is training new people, not knowing what the field is, the question is if it is something a developer could specialize in. the answer could be "no". for all we know the specialization could be in bee-keeping.
so for that matter, if anyone has recommendations for a software developer on what to specialize in that they do not want to share publicly, please contact me.
I heavily disagree.
To preface - I have a recent and a successful PhD in Comp Sci. During grad school I've known active grads who jump onto the "next big thing". Well, long story short - almost every "next big thing" eventually dies down, every single one of them ended up being a generic Software Engineer after heavily specializing in the "next big thing" because, unfortunately, the cycle of the next big thing just keeps on rolling and within 3-5 years the current "big thing" may become entirely irrelevant, which ended up being the case for them.
I understand this might not be very helpful, but I suggest doing the footwork to figure out the intersection of something you truly have an interest in as well as a tech with a potential and specialize in that.
Well they’re all employed and given they chased hype they probably are in a tech hub and therefore I’d expect them to be earning above median for their cohort and still working on interesting stuff. How many of these assumptions are wrong?
when i was young, being in a niche area allowed me to get jobs on 3 different continents and travel too. but eventually, that was not enough. i had to leave that niche and accept a python job so that i could work and live in china. it was great while it lasted, but in the long term it proved to limiting because the jobs weren't where i wanted to be (if there were any jobs at all. i stopped looking)
the problem is of course to figure out what has potential, and what it takes to get into that field.
It’s a trap to be scared and try to gamble on near future.
By the time most of today’s wannabe AI gurus are good enough to get a decent job offer - AI would not be such a big deal (best case scenario imo).
I didn’t notice any “obvious” changes in the game that weren’t there before.
For instance I don’t see anything different in requirements and comments of HRs reaching me.
JVM is still same jvm, bugs are still bugs, unit tests are still unit tests.
The fact that now instead of googling a problem you google and gpt it is a minor change in any real world complex engineering context that im aware of.
I agree on the AI stuff ... personally, I think it's going to need 5 years or so just to iron out all the legal and regulatory stuff before the 'next big push'
In general, I think you make your money in Tech at the ends.. Either bleeding edge or trailing edge. ATM, I'm working on Rust, serverless, playing with Pijul, keeping an eye on OpenTF, and pulumi.
For trailing edge stuff - I figure I'll be the worlds oldest COBOL coder by the time I retire and can use that to fund my retirement :-P
There are certain aspects of computing that will probably always be around. We'll always need fast compilers that generate fast code -- people don't want slow programs, and programmers don't want to wait for compiles to finish. Relational databases aren't disappearing anytime soon. And as we keep getting smaller and smaller devices, we'll keep shrinking things down, so we'll need new ways of representing and compressing data. The list goes on and on. If you can get comfortable with things like these, you'll always be useful to somebody.
(And yes, COBOL will be with us for many years to come. In the next century, we'll probably be emulating IBM mainframes on quantum computers.)
I think AI will still be important, but the AI of the future won't look like the AI that we have today. People have a tendency to redefine "AI" to mean whatever the new hotness is -- LLMs today, deep neural nets a few years ago -- but new things are created all the time, and fads always change. When I started grad school for comp sci nearly 20 years ago, I worked with a group that was big on "AI", but we were dealing with multi-agent systems. No neural nets or super-intelligences or anything like that. Neural nets were actually considered passé back then; I vividly recall a professor telling me that SVMs (support vector machines) were stronger than neural nets, because SVMs had a stronger theoretical foundation and were more amenable to mathematical analysis. Neural nets, on the other hand, just happen to work -- but they happen to work very very well! Deep learning didn't gain traction until after I had finished grad school.
The LLMs that we have today are amazing, but there is still plenty of room for improvement. Having to train it on a huge dataset is problematic for some uses; perhaps there is a related structure that can be trained more easily and more quickly. That would also reduce the effect of OpenAI's monopoly. LLMs also have specific weaknesses, like poor performance at arithmetic. At this stage, I wouldn't really feel comfortable feeding problems into an LLM and presenting users with the LLM's answers. It's still the Wild West in many aspects. There is always an improvement on the horizon, but it's hard to tell where it will come from and when it will come. Maybe we'll have LLMs that really start to resemble intelligence, or maybe we'll have a totally different structure that does everything LLMs can do plus more.
The obvious thing is gpt-6 (or even gpt-5) will be able to do your job better and faster than you. This will happen in 1-2 years. That doesn’t scare you?
I am bearish on that prediction. I’m not convinced LLMs can scale that much more. Of course, I’d be happy to be wrong. In that case, I’ll leverage the tools to become more productive.
Knowing how to code doesn't make you a Software Engineer in the same way being able to read and write in English doesn't make you an author. If you truly believe the current and future iterations of GPT can code better than you, look to elevate your own skill.
If ever, a GPT-like LLM/AGI ever exists where it can distill business requirements, understand modular designs and intelligently establish complex relationships between different systems and contexts, then 99% of all jobs will perish.
This will be an unprecedented disruption at a macro scale that humanity has never seen before. All of our current economic models will be instantly trash. How likely do you think this will happen? If it does, there's no real incentive to have these systems produce anything because no one can afford to buy anything. A global revolution would be inevitable.
unprecedented disruption at a macro scale that humanity has never seen before
Exactly.
How likely do you think this will happen?
I estimate the chance of this happening after the release of GPT-5: 50%.
Chance of this happening after GPT-6 is released: 90%.
You painted a pretty bleak picture of what might happen. I tend to agree with your assessment, but there's a chance that the total automation of everything will make things so cheap that quality of life will actually improve for majority of humankind. In any case, we still have a few years to prepare for this disruption.
p.s. I'm not sure why people are talking about software engineering as some sort of a subtle skill that is hard to automate. It's quite likely that GPT-5 will be able to talk to my boss, colleagues, and any other other stakeholders to understand what needs to be done, access all relevant information, iterate on requirements, ask for help, and problem solve, and it will do so better than me, or anyone else in this thread. Even if GPT-5 is not quite there yet, it's almost guaranteed that GPT-6 will be (barring some unlikely global catastrophic events preventing further technological progress). At that point, there will be no reason for my boss to employ me, or for his boss to employ him, and so on.
let's bet 1000USD that in 2 years there will not be even 2 confirmed real cases where any algorithm is doing a full time job of senior+ software engineer?
In two years the job of a senior software engineer will be completely redefined, so I'm not sure how we could test your prediction. Would someone who tells an AI model to build something still be an engineer, or would we call that person a manager?
including current versions quality degradation, unprofitable business models (it's still too damn expensive to be covered by your average 19$/m subscription) etc.
We simply don't know what it will be. Growth in nature does not continue at the same pace forever, and eventually the speed of innovation will taper off.
It's a common phenomena, described with an S-curve: (https://en.wikipedia.org/wiki/Sigmoid_function). We can't know if LLMs are already there, or not yet, we'll only know after the fact.
OK, so it is reasonable to assume that the chance the progress will continue in the next two years is about 50%, would you agree? If so, why are people in this thread dismiss my concerns as unreasonable?
I am a devops engineer working mostly with platform engineering managing over 80 eks clusters, I am learning more AI an embracing AI, it makes my job much easier and helps me write tooling faster. I notice people at work that resists AI and like to do things the way they have been doing it for years. I wrote an Azure wrapper that helps my team debug kuberntes issues in real time using AI , it also analyzes deployments and makes recommendations and this has caught the attention of teams that are now using my tool to debugg their workloads.
My company already is doing a hiring freeze and we have a lot of work so having tools like OpenAI has been invaluable to help me with my daily work.
> I wrote an Azure wrapper that helps my team debug kuberntes issues in real time using AI , it also analyzes deployments and makes recommendations and this has caught the attention of teams that are now using my tool to debugg their workloads.
I don't think it's obvious that the game has changed at all. I plan on doing what I've been doing to stay relevant in my career so far: keeping my skills fresh, tackling new problems, etc.
Not writing JavaScript for money ever again. I switched careers. Now all future jobs will either be in management or require certifications or someone else more naive can have it.
My solution in general terms is to go where competence isn’t ambiguous.
Or ubiquitous. When you are selling a skill that is as common as the ability to write JavaScript, then you aren't selling much, and you're very replaceable.
Learning a very niche skill is harder. Finding jobs with very niche skills is harder. But once you've done that you become very hard yo replace. (Your job might go away, but there's seldom outright replacement. )
Agree on the principle - but management is another very replaceable skill.
It may protect you from offshoring because the stakeholders don't want to talk with people in third world countries and want to zoom with someone in a nice office paying 2k rent per month for a studio.
It won't save you from AI. We jokingly built a chatbot to write in the style of our product manager and engineering manager and it's shockingly accurate, especially if you consider both figures ask what the stakeholders want and technical feedback to engineers who knows what they're doing and meet the requirements.
I've been asking myself this question a lot, and I don't have a perfect answer, but here is what I am seeing so far:
We are moving from a period of time in which engineers were needed to do, essentially, day-to-day grunt work of software development (write this CRUD app, figure out this schema, implement these requirements) to a period of time where engineers will be needed to oversee, design, and manage relatively intelligent tooling that will do those things for us, and then be evaluated on its results.
Put another way: Engineers are currently like factory employees at the turn of the 20th century. Lots of manual tasks are needed to keep our "factories" running, tasks that, in the 21st century, robots can do just as well. But that doesn't mean no humans work in factories. Plenty of people do, but what they do are the things that the machines can't be trusted to do alone, or at least, can't be trusted to do alone sufficiently reliably for reasonable cost.
But even so, far fewer people work in factories now (as a proportion of the population) than did in the early days of the industrial revolution. It seems to me that engineering will likewise be winnowed down. That means that ultimately even the most valuable engineers won't be as valuable. You won't need as many of them to do the work, and you won't need to pay them as well.
If I choose to stay in engineering (which is by no means a guarantee), I think I will need to focus on moving from "day-to-day implementation" into "designing and monitoring the overall approach to systems." At most organizations, this means getting to and being successful in, at minimum, a staff engineering position, preferably higher (e. g. lead/principal). I am nearly there at my current organization, but I don't have the skills to perform at the next level yet. I can probably develop them, but that's also not for certain, and even if I do, I might not like that kind of work.
In that case, if I wish to remain in the workforce I will need to change career fields, and find one of the things that won't be automated away by LLMs or similar technology over the next 15-20 years. (For example, contrary to a lot of thinking currently, I think a premium will continue to be placed on genuine human creativity; I don't think AI will eliminate the desire for humans to consume art created by other humans. Any field which involves physically doing - such as the trades, or maybe some kinds of hardware engineering - would also be an okay bet.)
Or I could always coast on the coattails of my spouse, who is already in such a field. That might be easier :)
I keep an eye on the trends and kinda dip my toe in a bit so that I know the basics.
Like with AI, I've tried Copilot (in the free beta). I've poked around GPT both on the web and using the API. I've tried local LLM models and StableFusion.
Not going to go all-in on any of that, but I kinda know what's going on and where the tech is at.
Now if someone pays me to take a deeper interest, I already know where to start looking and who to ask for more information.
Also: look for the boring stuff. There's a ton of work in the uncool industries.
There's software in tractors, mining equipment, smart metering etc. It's not the cool whiz-bang stuff where you can rewrite everything in Rust, but on the other hand it's stable work where the software is just a part of the bigger machine and you're usually not writing code with a fire under your ass. Nobody expects 16 hour work days and there's actual work-life balance.
Stay employable? I have been unemployed for 1.5 years. Absolutely NOBODY needs a product management leader anymore. If anything only entry level and I am overqualified to do that (not saying it arrogantly, but companies don't ever consider me stepping 3-4 levels back).
Tech is in a slow death spiral.
I've been navigating this question myself - in the mechanical engineering world.
Much of my background is in product design, particularly sheetmetal enclosures. I grinded through a few different jobs in that space, picking up some skills in CAD Administration, PDM (think engineering document workflows), and programming (mostly VBA and Python). I saw myself moving away from product design/manufacturing into engineering automation... but I kept getting pulled back into designing widgets.
In my early career, designing widgets was fun and interesting. Even got some patents. But indecisive customers and overpromising salescritters made the experience extremely frustrating. Some people just love to quibble with designers over minutiae, resulting in delays and overruns. "Sure, the product is functional, but... it doesn't look like what we had in mind."
I got laid off this past summer, but lucked out with two competing employment offers. Company A needed a product designer, and they needed a CAD/PDM Administrator as well. Company B needed someone to reverse engineer existing CAD models and program mesh generation scripts for simulation software verification models. In comparing these, I realized I had an opportunity to pivot away from designing new manufactured products - and negotiate for more money in the process. So I picked Company B, even though my skills were probably a better fit for Company A.
Since starting my new job, I've found myself lately enjoying lots of downtime. In product development, I was always scrambling to produce and change designs to meet ever-evolving requirements. The design process often trends toward instability and complexity as requirements contradictions arise. In contrast, reverse-engineering naturally reaches a point of stability, and quickly. It's very satisfying doing well-defined modeling tasks and knowing that the tasks are DONE.
Future-proofing a tech career is often seen as an exercise in skill-building, focusing on what employers are trying to acquire in new hires. But I think tech workers really need to think about eliminating their career pain points as well. Reducing burnout can open your mind to new career possibilities.
I'm creating a low-code platform using AI to automate my own work (to be more productive) and also trying to learn how to engage an audience on social media (around AI and software development) to use as leverage to get contracts, launch apps and make a profit.
As a coder since 2002 (when I was 12 years old), I feel like I got into tech at the right time. Nowadays the new hacker is the people that can amass a big audience doing interesting things. If I was 12 years old now, I probably wouldn't hack my way coding, I would do it through social media.
I am doing the same thing; this will be my 5th since the mid 80s with the AI (obviously) being a lot better than in the past decades. It's hard but fun.
I'm creating it from zero, creating a framework using Lit (moving from React professionally), PWA technology and LLM integration (to generate code using the framework)
Back in school in 2004, I had to make a choice: go into quantitative finance, the stuff was piping hot, and people there made a lot of money OR follow my interest in AI/neural nets, possibly not making much.
I chose finance, then the crises kept hitting. I don't regret it, because I'm still very employable and I did make some money. But to me it's a bit of a missed opportunity.
The point is: don't look at what is hype now, look at what could be good in a few years. My guess is that the next huge thing is robotics.
Definitely adding generative AI clients on my portfolio, then just using the latest tech I like.
I don't take side gigs unless they're AI, Rust or solid.js.
The focus is still on making products and live off them though, more than staying employable.
Staying employable is a nice side effect of training the "building products" muscle.
I don't think it will be too soon but at some point work done by humans will lose any meaning and we'll just divide in people with income/value generating assets and people without.
What can GPT-6 do for real? Just write single-step code? Whoever is writing that prompt will be a low-skilled person to fix their problem or find out whether they are doing right. No GPT will replace developers. Writers, maybe.
I've been avoiding having a real job for most of my adult life. I am trying to build an online business or two that leverages AI. Such as for fine tuning LLMs or an open source agent hosting system.
Although right now I don't have the energy to work on that stuff since I am scrambling to get the next crap Upwork contract.
If you are following the trend of the moment, your chance of success is low.
Everyone and their friend is "working on an AI business" right now. Some might make it. You might be one of them. Most will turn into nothing.
Again, with respect, jumping into AI now shows a lack of imagination. If you want to make something of value, you need to get outside the mainstream, be more creative, not something that's just riding the current hype wave.
The fact that you had to claim it was respectful multiple times indicates that the opposite is true.
I kind of asked to be disrespected by admitting that I am not having an easy life.
I got into generative AI awhile back when GPT-3 came out. A very significant portion of Silicon Valley also switched to AI related efforts. Not from a lack of imagination but because it was painfully obvious how powerful it was.
It's been enough to get by. It's definitely easier to get jobs on Upwork in a trending field.
Since I've been pursuing online business for over a decade and built many for others, I am quite aware of the odds of success.
What you actually informed me about was only your lack of respect.
None of the rest of it was even a tiny bit informative.
Perhaps to are projecting this as a way to defend against your poor ability to adapt.
I’m taking some time to go back to school and finish my bachelors. I work with a lot of rural ISPs so I think for the moment I’m ok. I might pick up some networking skills. It seems like networking changes slower than server side stuff cause it more often requires hardware upgrades.
Not sure about your capability but what I see in most of new devs and techs is serious lack of fundamentals. I would start there. Once you tick that box it doesn't really matter what's next. You'll pick it up quickly.
I used to believe this too. It’s completely true, but only a half truth.
It doesn’t matter how awesome you are, because in most cases employers are viewing the world as a bell curve and targeting the large population in the center. Being supremely awesome in your skills and capabilities moves you far away from the center of the bell curve.
Most employers realize many of the people in that swollen center seriously lack the fundamentals and adapt the environment accordingly, such as lowering barriers of entry and lowering minimally acceptable quality criteria. That is not the environment a top skilled developer will strive in.
My goal moving forward is to work in jobs with higher barriers of entry, preferably some manner of institutional barriers.
I work for a company that I think will last, and is good to work for so just hold on tight! I did move over to devops (platform if you prefer) which I think is more evergreen. So knowing both app development, being the person who can configure tailwind on npm, how to do backend queries and set up an ingress controller in kubernetes is kinda a good jack of all trades to have!
trying to get my Amazon AWS Machine Learning Cert. Certs definitely give you credibility in this game nowadays. Google Cloud certs are also good, as well as CyberSecurity stuff
Dunno, I got AWS CSA-P and it never got me any new interviews. Granted I stopped looking and only took interviews from people that looked for me actively.
