I need a few decades of at least mild security to get to retirement...
Are the trades safe?
Im worried about homeowners and companies ending up with robots that do all those tasks. Not next year obviously, but it could happen before the career change fully pays off
And if not the trades, then what is a better bet?
Don’t leave software. AI will change your job, but it won’t replace you, not even in multiple decades.
Be strategic about the kinds of software roles you get into. Look for niches with stability. Look to go toward highly regulated software niches. For example, I work in an area where it would be impossible to remove manual testers due to certain regulations.
Software work for government where you must be a US citizen or you must do specific things manually could be another example.
I also think some areas are going to be more insulated from AI disruption. I think ops and security have more stability. Anything that requires more judgment and less “take inputs and turn them into outputs.”
Moving from an upper middle class career track to something you can get into without a high school diploma is a mistake. In many trades you’re competing with dudes who are working under the table and aren’t even documented. Even if you’re not, they’re mostly hard on your body, and some aren’t even immune from automation.
It’s making a lot of progress in “correctness” and a general ability to accomplish complex tasks.
It’s making damn near zero progress in judgment and generally being trustworthy to be left to its own devices.
Software is still on the high end of the spectrum of career cushiness, it's just no longer a full on outlier. If you've already put a decade or more into this career track, switching to another career will have you starting at the bottom of a ladder that is significantly more rigid and inflexible than you remember the early years of your career being, with higher-run perks still no better than what we have here now.
Better to make the lateral move from 2010s-software-eng to post-AI-software-eng than to start from scratch at the bottom of an equal or worse ladder.
AI is used massively to identify vulnerabilities because it is not difficult for an AI.
Investigations or case management is marginally helped by AI, this will change with time certainly but for no it's a lot of hand waving.
As for AI in the software field… I’ve watched the people around me. Some of them are doing good work with AI. And what they all have in common is serious skill and experience predating AI. What it looks like from my seat is that AI in software dev and related fields is absolutely a force multiplier… but only for people who already know what they’re doing.
So as a few other people have said, don’t be too hasty to jump ship. There may yet be some good opportunities.
Btw, having said all of that, my AI use is limited to summarizing searches and assisting on code review. I’m personally quite “anti AI” and close enough to retirement that if eventually I can’t keep up without going all-in, I’m ok with being fired. Love my job and the people I work with, though.
It’s definitely an exciting time to be in robotics, though it looks like the Chinese are pushing more aggressively into it thsn western developed nations. AI is also a huge risk to the trades (the joke about the plumber just needing to know hit im a call means those quick calls are at risk because that knowledge is easy to get with a chat bot).
In particular, large discontinuities in economic conditions, which we just experienced recently post-covid, always produce downturns in affected industries (which is distinct from the first order aspect of how expensive money is). So I think even ignoring AI we could expect several years of down-ness and reorganization followed by a recovery.
I don't think we can even tell this point if AI will reduce demand for developers. The economy for software might be even worse right now if the AI developments hadn't happened when they did, there was an insane hiring spree during COVID and even before that the industry employment-side was growing at a probably unsustainable pace.
https://qz.com/tech-layoffs-tax-code-trump-section-174-micro...
Luckily the rule still applies to GPUs. Buy those and you get to write them off against tax. Even in Europe. Even in France if that is even possible.
There was this really weird moment where companies all looked at 2022's financials, saw the effect of Section 174 and the end of ZIRP, and started laying off software departments en masse, yet the media got all hyped up about how AI—at that time just GitHub Copilot!—was ending software engineering. It was absurd.
They did finally fix Section 174 again, but ZIRP is still gone, and it's important to remember that apparently the original autocomplete Copilot was enough to get the media to declare the end of software engineering.
If the doomers are right and AI fully automates knowledge work leaving most unable to produce much value, the entire notion of a 'safe career path' doesn’t make sense. Not just because of robotics, but also because demand for even the safest bets likely collapses in such a scenario. And honestly, in this case there would be far more pressing issues than long term job security.
IMO, make decisions like these based on what fulfils you rather than out of fear.
If all knowledge work disappears, you'll live in an economy that's so unstable from high unemployment that it will hardly matter whether your specific chosen profession isn't automated - on the mild end it will affect the demand for what you do, the cost of everything you need to live, your retirement savings, how you get taxed and what your taxes are spent on; on the more extreme end, it won't matter that being a plumber gets you nice house if there people outside throwing bricks through the windows.
Even useful businesses wont be able to find paying customers because everyone is too broke to pay
Technically, a massive demand curve shift downward due to only rich people having any money (and as a result, even they lose a lot of their wealth?)
One of the things that's recently been on my mind is that everyone is focused on the bottleneck right now, on "how do I make it through the next 3-10 years as we transition into this crazy post-scarcity techno dys/utopia.
I want to focus on what happens afterwards. I want to live in the Culture. What decisions should we be making now on this side of the bottleneck to make sure we're going to get the Culture on the other side of the bottleneck? That's what I want to be thinking about. I can't be the only one worried about this either?
Like, I'm building my little solar punk cabin, and have aspirations to travel widely, and slowly (by foot or bike or hell, maybe post-singularity zeppelin!), and explore the world, and explore myself, and really truly live.
I can see what that world would be like - I can feel it even. But... how do we get from here to there? That doesn't seem to be a question anyone wants to answer? It's all just "what do I do for money?!" and "what's your P(DOOOOOM)" and not actually talking about what kind of world we actually want to live in.
Plummer is probably the safest of the trades if I had to guess. That said, I'd encourage you to think more about other things humans are likely to always want/need - see if any of the other options are interesting to you? For example, lots of people are and will continue to be lonely, having a bar to sit at with someone friendly/interesting is a tool we use for this (I wrote a blog post about this idea: https://b.h4x.zip/finale/) - but there are also things like undertakers, kids activities centers/camp councilor, etc. Not all of these things will be equally financially rewarding, but maybe you're able to think about rewarding in other ways. :)
[1]https://www.youtube.com/watch?v=4whgw2gLBS8
[2]https://www.youtube.com/watch?v=C5MsqzQW8X4
[3]https://www.youtube.com/watch?v=fhK28R9J-DQ
[4]https://nico-bohlinger.github.io/one_policy_to_run_them_all_...
[5]https://www.pi.website/blog/api-beta
Trades are high floor, low ceiling. Initial wages are good, but after journeyman there isn’t much growth. Unless you start your own business, then sky’s the limit.
Trades are hard on the body, there’s very little work-life balance, and they are very boom/must cycle dependent.
Skilled tradesmen who can mix business acumen are rare, thats where I’d aim for.
But for wanting just a “job” software is better and more comfortable. Can always do woodworking in your spare time too.
Companies can 3-5x each of their employees’ output and yet they can’t come up with products or lack a backlog to make them useful, so they let them go.
For some it went well, for others the think they jumped too has died already. Some discovered that having main-stream skills just makes you a small fish in an enormous pond.
Which is to say that predicting the future is hard. And predicting what skills will be most valuable 30 years from now is pretty impossible.
Yes, I know the present is scary, especially if you are just starting out. All you can do is be adaptable. It's unlikely the tasks you are accomplishing today will be what you are doing 5 years from now.
The running joke is that they're called "the trades" because you trade your health for money. Not trying to discourage you, but be mindful of this. Also they're often accessible only with proper licences, which take time to acquire.
My personal plan B is learning an uncommon foreign language. I'm not a native English speaker and yet I'm here, writing this comment, so to me it's just a question of putting some time into learning.
AI is great at context-aware translation in writing. But if you want to actually communicate, better really learn the language.
You will probably be safe from the sort of automation around the corner as the boom is largely contained in the virtual world. There is no such thing as job security in a dynamic economy like ours. The ones who are the most successful are the ones who can stick and move and are unsentimental about what work they are doing.
AI lab offers free plumbing and HVAC repairs across three cities to train AI robots
https://www.tomshardware.com/tech-industry/artificial-intell...
It’s not like white collar work like medicine or finance isn’t some stress-free ride. Software work is unique in its potential to sit around all day and do nothing, but I still regularly do 10+ hour days that are mentally exhausting. I’d prefer physical exhaustion personally, at least my body wouldn’t be metabolically useless.
On a side note I’ve never met anyone in construction that envys my desk job. They generally like moving about and having a lot of functional strength.
One desk job I had involved helping the young company finish an office; including putting up studs, running cable, drywall, mudding, painting, etc. I've done some woodwork for home projects too.
Manual labor is certainly not for everyone. Many of us just are not built for it, and our bodies will suffer greatly trying to keep up.
I was once in the trades (carpentry). Realized 2 years in it's untenable to make a living in the bay area and eventually found my way to software.
If you get paid a lot, it's hard to pass interviews.
If it looks easy, there's some hidden drawback.
If it seems like it will always be needed, someone is working on a replacement.
If it looks well protected, it's from new entrants.
The money is in high-end customers, and one's ability to sell the narrative and be reliable in that capacity. For example there is a ton of money to be made running a painting business, but very little in being a painter. Same for doing custom cabinetry. My personal favorite was always tile, but I also recently replaced a red oak floor down to the studs based only on what the sales guy told me how and my god that came out beautifully. Studs up to the sanding to the Swedish finish was all me. He said most floor refinishers are booked out 18 months and the money is very good. Urban PNW.
Trades delineate by culture / ethnicity but in my last foray into the trades I noticed most of the workers are now native Spanish speakers. Customer-facing skews White.
Most people with money have no idea how things work, and they're afraid and not inclined to learn. They want someone they can relate to and "trust".
I've eyeballed going that direction but frankly it just scares me to be self-employed.
Like in Tech, the good Tradies are already doing what they enjoy. So if you haven't done drywall 'just because' it's not a field for you.
I've met a lot more 9 fingered mechanics than software developers.
>jumping ship
that sure seems unwise to throw away a position in a time of pre-supposed insecurity.
Most trades are small businesses, relatively badly managed, and pretty incompetent outside of their core competencies. You could add value there.
I imagine in passenger planes, two pilots would be mandated for quite a while, even if they don’t do much. Planes can fly on autopilot and can now land/take off automatically, but I don’t see people getting on a plane with no pilot.
Once the Russian and US wars wrap up - there’ll be more demand for flights.
Freight flights are over though, drones will take over there.
I highly recommend the career field in general. Especially small airplanes and becoming a mechanic. Robots will be able to do the work of an A&P mechanic eventually, but someone is still going to have to be legally responsible for the sign off work.
But to your point, you might not get a choice.
I have a "trade" hobby (Aircraft Mechanic) that I've been leaning on harder as my burnout in software increases -- and my tolerance for the toxic techbro culture goes to zero. It wouldn't work if I had just chosen that at random, I enjoy it and derive immense satisfaction from it. I'm oddly lucky that at my tenure and experience level, I can pull tech money from it.
So if you don't have such a tinkering/trade-adjacent hobby, I think you should prioritize cultivating one and see if you can fix that dishwasher, shower, raingutter, or electrical outlet. Be safe and sensible about it. Maybe one of those things seems more "your jam", then, yes, why not fan those flames and start getting certifications/licensures/experience on the side? Seems like a perfectly sensible thing to do in uncertain times to me. Worst case, you invested time and now never pay an electrician/plumber/handyman again?
$0.02 :)
If you just do this 10-20 hours/week, after a decade or so you can become very skilled, have a large following on YouTube, and if the worst happens to your software career you can easily step into furniture making full time. Even after 5 years you’re going to be pretty good at it with a decent following.
This also makes it much easier to pick up other trades because you’ll be practicing similar skills.
Many people die young and fat and very sick at desk jobs. Inertness does not seem to equal health. A job like industrial maint electrician involves just enough walking around to outlive the average desk jockey. Of course there are very fat and sick heavy equipment operators...
Another meme response you'll get implies its all your choice to "stay in" whereas the problem is the market is contracting so fast you won't be allowed in if you're not already in, and if you're kicked out (layoff, just not keeping on the official mandatory career path, etc) you'll never ever get back in. Don't worry about deciding to leave software development, plan ahead for when you're uncontrollably permanently kicked out. Its not going to be your decision.
Education can be helpful on the jobsite. I've seen fine finish woodworkers do very weird, yet successful, graphical solutions to avoid five seconds of trigonometry. Automation PLC programmers will do weird stuff to avoid bode plots and similar control theory topics. There are people tuning PID controllers on the job who don't know or understand the P, I or D, they just twiddle the knobs until it works right, often while developing all sorts of magical mythology.
Keep in mind the wide spectrum of the size of your counterparty. Working for a multinational megacorporation at a mine operating heavy equipment for your entire career will be wildly different experience from working for a single human individual digging a septic field for a day.
An interesting analogy with lawyers is people like to think their interface is like a chatbot so a chatbot will replace them. But people actually HIRE lawyers because they want to outsource the risks of doing something legal-related to someone else, in a way its an insurance gig not a chatbot gig. You'll find in the trades they're not really hiring a guy to move the levers to dig a trench, they're outsourcing the risk and hassle of owning your own excavator machine and the risks of screwed up one-call utility marking, etc. In a similar way a robot arm that can screw a wire to a wall switch won't replace an electrician who knows how to run wires without taking an axe to every piece of plaster in the room. Its fascinating how the invention of CNC machines has not resulted in machine shops closing down and every company broom closet having their own personal robot mill and lathe, but it has changed the entire experience of working at a machine shop... See also 3d printing LOL.
Regardless of the industry, wage slavery is insecure by design . The threat of losing your job is ever present and used as a motivating factor. Even when it is not being overtly stated, it remains an option that is never far away from being utilized.
Real security is control over your own destiny. And the only way I know to achieve a reasonable measure of it by running your own business. It doesn't really matter what kind of business but I'd recommend something you have some natural interest in.
Just as an example, I know a guy with lots of financial security. He owns a string of barber shops and hair salons. He started out by learning to cut hair but nowadays, the only thing he cuts is checks to his employees --- while taking a sizeable "cut" for himself. If AI robots learn how to cut hair, he has the money to invest in some, adapt and continue on with his business. This is called "security".
No one ever rushes them and everyone's happy to see them. Plenty of overtime.
I got the feeling that it's pretty hard work and long hours and of course the danger of explosion. But somehow the crew was happy.
Its a bit like saying, this administrative job is unstable, i'm going to buy an instrument, learn to play music, and become a concert soloist, because there will always be people who want to go to a show.
its interesting to think about what safe means though. nobody was taking risk in a software dev job, its basically a rarefied trade. its the opposite of risk. there is no speculative value, no significant opportunity cost, no leverage, no negative sum failure mode, and a until recently a high degree of certainty about finding work.
evem sw startups by young people are incredibly low risk until they have to make payroll.
by comparison, professions are a bet they can navigate something where, after they have the credentials, it is an undifferentiated narrative war of all against all without any consistent code of honor or objective standard of competence. "political" as it were. there is no safety other than the credential.
software is the safest choice after maybe accounting and dentistry.
Trades continue to boom.
Personally, I'd pick electrician.
Plumbers also do very well, but in addition to dealing with the water flowing in, you have to deal with the shit flowing out 8-/
A lot of modern electrical installations are on a level of an advanced engineering technician.
The work isn't as physically taxing as other trades, but it is still physical. Pulling conduit is hard work.
But the pay is good, and it'll leave you fit. You'll probably live longer...
I don't know why everyone thinks knowledge work is going to implode and only "the trades" are going to exist. How can all those people that lost their jobs to AI afford to hire a plumber to fix their toilet? How can the property developers afford to hire carpenters when all those knowledge workers lost their jobs and can't afford housing?
Society just doesn't work this way, and likewise no one wants there to be some kind of imminent collapse. It's far more likely you'll get hired to do something soul crushingly boring or useless in the future compared to the wild and crazy idea of doom and gloom AI terminator fantasy. I can imagine many jobs becoming bullshit jobs. If you give someone a job, they can afford to buy things, and them buying things keeps the economy working.
It’s not easy work. Lots of potential for injury and seasonal boom and busts. So it depends on your current finances. Also have to be an apprentice first and finish the licensing exams.
Because you know software you can program these things, which are easily modified, to be jammer resistant meaning when noise floods the frequency they can either automatically change to a new one or raise the alarm as security is about advanced warning so someone doesn't surprise the owners.
Look up everything you can about surveillance mapping to identify vantage points or better yet go attend local training about this from any executive protection company in your city or online. This means where do you point some of the cameras. Someone looking to do crime on a business or detatched property is going to pick the spot on the streets surrounding the property that allows them to collect the most information on that target without being seen on visible cams such as a door cam. Walk around these properties and pretend you are the criminal, where would you camp out and never want to have your vehicle or face seen by the door cams meaning they won't be driving past the house or sitting directly out front. A pro robbery crew will use google street view with in-person surveillance to find the best spot to place their camouflaged wireless cam with battery pack which means they need to physically show up, identify the best vantage point and hide the cam in landscaping. YouTube has videos on detecting hostile wifi or persistent cell signals too and these devices are cheap to buy and write a custom program for.
Even unhinged maniacs will do some kind of hostile surveillance. An example is one of those copslop bodycam vids on YouTube about the guy who put up his house for sale that had a bunker inside full of food. A maniac saw that house listing and decided he needed the bunker to survive some unmedicated doomsday delusion and spent weeks surveilling without being detected.
Now since you are a developer make software that reviews the footage and alerts when a vehicle or person is at one of your identified vantage points. There's open source software for this or a simple screenshot will work and use chatgpt/claude to compare the baseline to the new screenshot at intervals you select. You can now produce a timestamp for the owner to manually review or decide how this should be done after training by some security/bodyguard company. They will also teach what can't be filmed because of local laws.
Take your security consultancy to a local farmers market as they want to be alerted to someone crossing their property too as a good IR beam scheme doesn't alert on animals.
A way to start is offer to install google nest (now Gemini for Home) or doorbell cams on Taskrabbit or local Meta marketplace and you meet them in person to sell wireless beams or your other security. If there's no subscriptions being peddled and you are a decent developer that knows about fault tolerance and quality management of product so no false alarms or weather destruction or failing batteries you could start a nice side business that may lead to full time. This equipment can be modified to detect flooding and fires too which is why pro security companies charge so much. Even if Gemini for Home totally eats up the home security space with AI there's consulting room in there to physically hook all your equipment into Gemini to control such as cams watching vantage points.
This is only one example where software expertise can be combined with a trade. People also want home automation still and tradesmen themselves use all kinds of salesforce software for CRM and billing which they would like customized but can only find sketchy remote developers instead of speaking to a local.
Robots can't take risk. We're transitioning to a risk based economy.
You're not hiring a XYZ to "do" something because everyone uses industrial production, robots, and LLMs to "do" things. You hire someone to make the headache and risk go away.
I need a trench dug in my yard. I own a shovel. I could own a robot to operate it or do it myself as healthy exercise. Doesn't matter, its just so cheap and easy to hire a guy to do everything from pulling permits to onecall utility marking to restoring the grounds when he's done. I can't afford not to hire the guy. I don't know or care if he operates the levers or a "robot" or a LLM, all I know is its cheaper to pay for a trench that to try to trench all by myself.
Another example. I "can" labor to replace my car thermostat in the winter, but the risk of breaking something or getting frostbite is high, the weather will never cooperate, its just generally a PITA, I hire my car mechanic to replace my thermostat in the winter. Car mechanics already only get like $12/hr they can't get any less, the remainder of the $150/hr shop rate is already spoken for, so robot car mechanics aren't competing at the $150/hr level they're competing at the $12/hr level. Good luck with that financially LOL.
On a nice sunny spring day at 65F when I need a mental health afternoon from work I'd do the task myself and enjoy the new experience. Almost all days are not like that LOL so outsource the risk and hire someone.
I don't have the capital resources to own every possible robot arm and gadget. There are many homes in the USA that don't have a "workbench" or own a screwdriver. The kind of people who currently can't be bothered with or trusted to own a screwdriver will never own a "service robot" to maintain their house. Those people will eternally hire skilled trades.
This is before you get into financial arguments where small players will always find it cheaper to rent obscure services. A dedicated sewer cleaning robot with LLM virtual reality interface, or just a dumb AC motor auger machine, either way very few people unclog their own drains and the few people who invest the capital into a solution tend to make bank... You'll always have a financialization scheme for making money as a small timer. How many companies occasionally need a pressure washer vs actually own a pressure washer...
There's not going to be solace anywhere eventually. Figure robotics, for instance. It's eventually game over for our current way of handling scarcity. Here's a Marx clip about this:
"Labour no longer appears so much to be included within the production process; rather, the human being comes to relate more as watchman and regulator to the production process itself… As soon as labour in the direct form has ceased to be the great well-spring of wealth, labour time ceases and must cease to be its measure.
Capitalism thus works towards its own dissolution as the form dominating production."
Once all the software engineers move into the trades, there will be a strong downward pressure on wages there, and also a strong incentive to automate there too.
> I need a few decades of at least mild security to get to retirement...
I hate to be the one to break this to you... but... there's no such thing as security. Like, I'm really trying not to be an ass here. This is my second career actually, I'm about a decade into it. I was a pilot through my 20s...
Then I got sick.
There's no such thing as security under the current paradigm. I had like $15k in the bank when I got sick - it was a total reset. I built back, went to grad school, etc. but yeah. Complete and total reset. So, out of the blue, anything can happen. Try not to worry about it too hard. Christ, if you're about my age (I'm late 30s), then I don't think we're going to get a real retirement. But I also don't think we're going to end up working that hard. My strategy thus far has been trying to build a cabin in the woods that's off grid with low property tax, that's the first plan, then see if I can limp along with my little business until the damn breaks. I would absolutely love to do something more interesting, technical, that required some more schooling, I would love to go get my PhD. I would love to live in Spain for half the year - or full time. But... getting the right remote job, or building out my business such that I could make that work is much much harder / has more moving pieces than you would think.
After Riemann, I'm not sure what to do now. If they can solve Riemann with Claude, then it's only a matter of time before robot control is cooked too.
> Are the trades safe?
No.
> Im worried about homeowners and companies ending up with robots that do all those tasks. Not next year obviously, but it could happen before the career change fully pays off
I think we'll start to see robots in the wild in the big tech hubs by the end of '27, and probably start to see them literally everywhere by some time in '28? This presupposes there's no cataclysm or or apocalyptic events that occur in the mean time? But yeah. There's a company called Nori that wants to sell a $1700 robot. I keep reaching out to them though to ask questions and never hear anything back? So it's plausible they're vaporware? But yeah... it's coming.
> And if not the trades, then what is a better bet?
The only strategy that I have been able to really figure out has been, "work for yourself" and "try to get into something that has some regulatory hurdles." Engineering maybe? Other fields too like medicine?
But literally in the last few days I've been starting to mull over the idea that maybe looking for moats is stupid? Maybe the "killer" education to get is broad-based general education that touches many disciplines? Or, if you want to be doing something technical, something like physics or chemistry, etc. where you get a broad based knowledge of how and why things work the way they do. Then, hypothetically you can understand what your AI agent is actually doing when you say, "hey, I want you to design this widget for me." At least then you'll understand the output of the machine.