Well said! 5 years ago I could have a pretty solid picture of what my career would look like through retirement.
Today, unclear what that would look like as well as unclear if the career will even exist that long.
One sunny spot I have is that life is very long. I think I have 1.5 careers left in me, and a full pivot is alright. I've been talking with older folks, and loads of people have historically done giant pivots at midlife.
> One sunny spot I have is that life is very long.
Not sure exactly where you are in it, but as a 53 year old I disagree. The younger I was the less likely I would have been to disagree but imo/ime the older you get the faster it feels like it is flying by.
> and a full pivot is alright
The AI-related existential dread is that it is very unclear what a pivot looks like when the ground under everything is shifting all at once.
The shifting is certainly a lot more obvious in our career than some others, but all of that shift (so far) happened more or less in a mere 10 months and is accelerating. And I don't think it is very likely that our career will stay anomalous for very long in terms of how much AI impact there is.
yeah, you're down to like 3/4 a career left, it makes sense that you feel that way. Less runway ought to feel different.
I do agree that a safe 'next' is unclear. But I also think if ai truly hits all knowledge work as hard as it might we're looking at up to 50% unemployment. Frankly there is no point career planning at all in that scenario. We will all have much larger problems.
And if it doesn't, then pivoting to some field with a combination of rewarding work and 'enough' pay seems sensible enough.
>some field with a combination of rewarding work and 'enough' pay seems sensible enough.
What makes you believe those jobs will still exist or will be available to you competing against massess of other now unemployed software devs? Overabundance of desperate workers will depress wages.
> if ai truly hits all knowledge work as hard as it might we're looking at up to 50% unemployment. Frankly there is no point career planning at all in that scenario. We will all have much larger problems.
If that much. I'm 52, and I wish I had decided 20 years ago that I'd rather do something meaningful. Like become a doctor. It would not necessarily be impossible to pivot to that now, but it would be very expensive and have a good chance of failure. If everything went smoothly, I could be a doctor just in time to retire.
> One sunny spot I have is that life is very long. I think I have 1.5 careers left in me, and a full pivot is alright.
I'm 41 and I could, in theory, pivot. But realistically: I'm good at exactly one thing in this world in terms of job ability. I can cook, I can bake, I can play music, but none nearly well enough to actually make a living on it. Thus the end of my tech career, if it should come to pass, basically means I will have to work menial jobs to get by. The future is very bleak for me, and the only real hope is that the tech industry will wake up to what a bad job LLMs do compared to humans. But while I think they will do that eventually, there's no telling how long it will take, and the thought that the market will eventually correct is cold comfort if I can't get a job in the meantime.
>loads of people have historically done giant pivots at midlife
Lots of people didn't have their field automated away though alongside most other knowledge work fields, or seen at the same time a drop in demand in manual work and an oversupply (drop in demand because of upcoming AI layoffs, and oversupply in manual fields, because of same)
Also I imagine a lot of demand for plumbers might drop because AI might help you to learn to solve easy cases. Imagine you turn on you camera and it would just tell you what to do exactly.
I have just called a plumber to do a leaking radiator replacement. Not because I don't have the tools or I don't have YouTube, but because I didn't want to do it.
But get a little more squeezed and with less salary due to AI eliminating your job or your field's salary bargaining power, and you'll try to not call one if you can avoid it, even if you don't want to do it yourself.
Currently AI made me way more work and unpaid over hours already buy sorting out the 10x expectations and the slop grenades by the 10x workers, that is why i needed it... so I can't even have this much feeling of having accomplished anything myself.
I'm not sure how this will work out. Burnout will be standard, and almost universal soon, apart from the previous low-performer, low skilled bottom of barrel folks who will mass produce crap with this.
Yes, buuut: I think we are already really close to that for a long time. You could learn a lot just from internet and youtube. But most people, I feel, just do not like to neither do plumbing/diy nor learn about it. Like if I tell people they can save hundreds of € by changing their own brake pads and that it is really easy to diy, they just shrug and go to the mechanic.
So I am not sure if AI will really increase the percentage of people willing to learn these skills and do repairs etc
Yeah the information has been available for awhile. What stops people is all the indoctrination and conditioning.
"I'm not allowed to replace my water heater, the government says I need permits, the government says only a plumber can connect a gas line"
I don't think people realize that the people doing entry level skilled work like replacing your water heater or swapping your brakes at jiffy lube often only exceed the caliber of labor that's in a McDonalds when it comes to attendance and pissing clean.
Some stuff you really shouldn’t do just because you saw some youtube videos, and it has fuck-all to do with “government indoctrination.”
I grew up in trades. I’ve seen shit work and dangerous work and serviceable work and beautiful work. And sometimes even I didn’t actually know the difference until an inspector (and family member) explained it.
So I do small stuff. If it’s serious and I can afford it, I use a trusted professional.
> So I do small stuff. If it’s serious and I can afford it, I use a trusted professional.
I'm with you here. Though I have to say, on the flip side, oftentimes when I choose to DIY something it is because I don't necessarily trust that I can find a professional who won't cut corners in ways not obvious to me. I've seen it happen too many times. I may be a lot slower, but I'm extremely thorough at the stuff that I do myself.
Plumbers are paid well because of the way they organized.
-They restricted the number of apprentices. That limited the number of people in the field similar to medical school.
-They could bar substandard workers from practising in the city of London.
-They kept their craft knowledge in house and it was never written down. Apprentices swore oaths never to reveal the “secret of the mystery.”
If you look at programming we do everything the opposite. It’s no wonder salaries will collapse. In fact the only question is why didn’t they collapse sooner.
Definitely. IMO, plumbing is one of those trades that videos on YouTube seem to help a lot more than an LLM. Code and simple stuff an LLM is great, but anything more complex or getting into heavy installation not so much.
pivot to what though, if "mind" jobs are going to be automated, it's not like there's a category of jobs that opens up, and going from mental to physical jobs in your 50s doesn't seem that easy
Physical jobs will drop in demand and increase in supply too. Unemployed "mind" workers don't have much to spend to drive much demand in manual work, and they might even try to pivot themselves to manual work, making things worse for manual workers.
and those who did the "mind jobs" 30-40 years ago retain that wealth, while wages plummet and they don't have to give up any significant part of their wealth to get what they want. This destroys class mobility , creating a type of feudalism.
Because we have a largely paper pushing service economy and so the guy who's plumbing a luxury condo is dependent upon the existence of some overpaid javascript monkey who the builder/financiers are counting on to buy the condo when it's done.
If you automate the paper pushing what happens? Jeff Bezos isn't gonna buy more condos.
It seems unlikely "physical AI" will be a complete bust but the post you're replying to is positing that fewer people will have money to spend on their services.
A lot of well-paid physical labor is going to face more competition from DiY + AI guidance among able-bodied and cost sensitive adults.
Even 1.5 years ago I replaced a blower motor in my vehicle with something like Sonnet 3.8 as the driver - saving myself $400 in labor and having a fun and educational experience to boot. Many home repair items fall into categories like this, where the actual physical skill demands are fairly low, but skills + background knowledge + supplies-on-hand were the value driver.
The BG knowledge is now available to anyone for free, and it comes with its own knowledgeable shopper.
Yeah, fewer "office jobs" and more automated jobs in general, means less demand for manual jobs from those (as money will be tighter).
But also, many existing office jobs type will try (or be forced) to get into manual jobs, and many young people will see AI-dominated fields as a dead end and also try to get into a manual job career, increasing supply, and further driving manual job wages down.
Exactly. There is no safe career. None. If AI replaces even a modest chunk of the knowledge workers, salaries across the entire workforce are going to tank. Some very specialized areas that require a lot of training may be insulated for a time, but even they will inevitably feel the impact.
I'm glad I'm not inclined to be a doomer, because the future would seem pretty bleak.
It's a spectrum, with "physical" and "mental" at each end.
Some of Software folk i know move to repairing electronic devices for living. He still works in AC-room, but instead of holding mouse and keyboard most of the time, now he holds meter probe and soldering iron.
this is probably a personal problem, but sometimes I get turned around on 'fun making a game' vs 'making a fun game' and I have the idea that ai elides this distinction even further
yes, this is a very helpful approach. I take a similar approach, partly because I find bringing up negative directions seem to bring the undesired action/direction into the agent's view and makes it more likely that they will go there
Deontology - `normative ethical theory that the morality of an action should be based on whether that action itself is right or wrong`
reply