Top
Best
New

Posted by poisonfountain 17 hours ago

LLMs are eroding my software engineering career and I don't know what to do(human-in-the-loop.bearblog.dev)
871 points | 858 commentspage 11
goodrun 16 hours ago|
I read all the posts in this thread - but no one has a good idea to avoid software developer obsolescence. My guess is this profession has 5 more years. It was a good run while it lasted.

All the other white collar workers are in the same boat. A pillar of the economy is going to be destroyed with no obvious replacement in sight.

cambaceres 15 hours ago||
Ok so it’s five years now? I’ve heard for the last three years that software developers will be out of job within the next 6-18 months. I’m glad to hear I’ll have a job for a little bit longer.
rootusrootus 15 hours ago|||
5 more years? That’s quite pessimistic, given how much evidence we have that LLM coding has as much of a long tail problem as any other tech we’ve created.
xyzal 16 hours ago||
Which other profession has the same amount of training data freely available for the taking?
goodrun 15 hours ago||
Don't need much training data for bank/insurance/retail analyst work - it's just basic reasoning and data retrieval. If AI could crack the programming nut - one of the most intellectually challenging professions - it can handle the rest with ease. The only human role will be high level monitoring - and even this will be largely automated so fewer will be required.
mschuster91 11 hours ago|||
> Don't need much training data for bank/insurance/retail analyst work - it's just basic reasoning and data retrieval.

For insurances... there's a reason why the three bullets of the plumber's brother were labelled "delay, deny and depose". You don't need a grunt to compose a denial order. Just let AI default-deny everything, most people won't have the energy left to battle the system or they'll die anyway before the claim finally sees an independent judge.

And as long as insurances aren't severely punished for denying claims that are found out to be valid later on, this dynamic will just continue as-is.

goodrun 15 hours ago|||
Sorry to reply to myself but I just recalled the recent Apple ad that ran last year about a corporate goofball who got his Apple device to fire off well crafted email slop to his manager who looked surprised/impressed. That trick will only last for a short time. The joke is that people like him and his manager will be the first to be fired.
lordmoma 13 hours ago||
looks like you just need a bit of harness for your AI: https://leestack.dev/writing/nasa-rules-for-code-that-cant-f...
jgilias 14 hours ago||
> And we all know the demand is drying up.

I don’t think the data really supports this? Last I checked at least.

skepticATX 16 hours ago||
The reason that I’m looking for an out is that it’s turned everyone I work with into imbeciles.

Nobody wants to think anymore. Coworkers are now just intermediaries for their LLMs. Talking to them is just talking to the LLM - sometimes directly copied and pasted, sometimes minimal effort to conceal what they’re doing. It is so disheartening.

And the sad part is, LLMs are incredible and can enable you to do much better work if you can stay in the loop, and stop focusing only on shipping speed. But from what I have observed, very few people care to do this. Who cares about substance when middle management thinks your productivity is 10x?

aogaili 15 hours ago||
Software engineers with low self-esteem who built their entire identity as mechanical cognitive workers are having an identity crisis and spreading FUD.

Currently, LLMs are nothing more than amplification tools that require significant steering. If you think your job is mainly to take input from POs or managers, translate it into if/else statements and loops, and review PRs, then you never really understood your role. Software engineering—for those who went to university and studied it—is fundamentally about complexity management and cognitive automation. People in the field, or at least those with some math background who studied software engineering properly, understand that it's all about managing complexity; current tools are nowhere near replacing a software engineer. What they call "taste" is imagination, creativity, embodiment, a more intuitive understanding of context, and yes, superior intelligence compared to current AI. However, AI and LLMs are excellent at mechanical work and mimicking human intelligence, so use them for what they are, and stop whining.

Going forward, the world is ever-growing in complexity, and automation will become widespread everywhere. LLMs just unlocked another level. So basically, cognitive work will be automated—perhaps up to 90%—until the next breakthrough (if ever). You can sit and cry, or you can learn the tools and help shape the future.

Software engineers can automate the entire economy now, including the executives, yet they just sit there whining and crying. This is a self-esteem, confidence, and identity issue more than anything else.

jplusequalt 15 hours ago|||
>You can sit and cry, or you can learn the tools and help shape the future.

What exactly are you helping shape? The volume of your employers bank account?

aogaili 15 hours ago||
Chinese Gen Zers are starting companies before graduating, people are generating music and starting their own studios, others are improving models and building harnesses, and the rest are on a mission to automate the entire knowledge economy—from healthcare to governance.

Regarding your employer's bank account: if that is all you were doing before, then that is all you will be doing after. You are just complaining about capitalism now. The irony, is that the means of production is now in the hands of millions. Those who are crying are those who paid their mortgages with for loops..well, I think they will continue doing so, with less hubris that's all. LLMs are nowhere near replacing full engineer.

So get a grip fellow engineers.

jplusequalt 15 hours ago||
>Chinese Gen Zers are starting companies before graduating, people are generating music and starting their own studios

Both of these have been happening before the advent of LLMs

>The irony, is that the means of production is now in the hands of millions

The "means of production" means jack shit unless you have the capital to scale up rapidly

>Those who are crying are those who paid their mortgages with for loops..well, I think they will continue doing so, with less hubris that's all.

Why is it hubris to give a damn about you spend 40 hours a week doing, or to lament change when it works against your enjoyment of those 40 hours a week. God forbid people value their time in any way that isn't monetary.

aogaili 15 hours ago|||
> Both of these have been happening before the advent of LLMs

I'm not sure about that. I read they are making better use of AI to accelerate building their businesses. Apparently, in China, people were not looking to work in corporations anyway, so they saw AI as a means to escape them.

> The "means of production" means jack shit unless you have the capital to scale up rapidly

There are people topping music charts without even having a brand; they just produce good music. There are people automating entire marketing pipelines to minimize capital expenditure, and there are people building niches for small crowds and making a good living out of it. Not everything needs scaling.

> Why is it hubris to give a damn about you spend 40 hours a week doing, or to lament change when it works against your enjoyment of those 40 hours a week. God forbid people value their time in any way that isn't monetary.

If you enjoy writing loops and if/else statements, you can still do it, but the market won't pay you when there is a tool that does it faster. That is the nature of the domain. Have you ever thought about the jobs that software engineers automated? What do you think those people did? They adapted, learned the tools, and moved on. This is the first time we are seeing automation at this scale in software engineering, and the reaction of software engineers is exactly the same as those in other fields.

Adapt.

aogaili 14 hours ago|||
Regarding hubris, I've been in this field for 20 years, and there are people in it who are just intolerable, frankly. They memorize every Vim command, refuse to use any other tools, and treat everyone else as less intelligent simply because they can write code... those people are getting humbled hard right now.
trumpdong 15 hours ago||
It doesn't matter to your boss. He will still fire you and replace you with a slop machine. Then you will not be able to get a job again and you will have low self-esteem.
aogaili 13 hours ago|||
That is limited thinking. People are replacing their bosses, and your boss will be bossing what exactly? Right now LLMs cant' replace the entire spectrum of human intelligence. But if your entire work is just translating your boss ask to for-loops & if-statements, the I guess yes, I would be worried.
trumpdong 7 hours ago||
Oh, my boss laid me off for an AI slop machine. How do I replace my boss?
aogaili 6 hours ago||
Build something? If the AI is that good then us it..and if it ain't that good then your boss did a mistake?
trumpdong 6 hours ago||
My boss did a mistake. How do I put food on the table?
aogaili 2 hours ago||
And what is your point? again, the tool is here to stay, it won't replace software engineers as of now, you can't undo progress, so what is your point exactly?

Same point as the textiles industry in the 18th century..and software engineers automating other industries for the last 50 years for what? for fun?

aogaili 13 hours ago|||
If you really think those layoffs are due to AI, then you haven't worked in corporations long enough. If companies are not hiring, and are firing, then what do we need middle managers for? In fact, if an engineer is really smart and masters the tools, what do you need POs, managers, executives, and even sales for? If you stretch things to the max, those who master the tools are positioned to automate everyone aside from the capitalist, as they are protected legally
trumpdong 7 hours ago||
You're thinking of this as a game of merit instead of a game of power accumulation that it actually is. Middle managers, like junk DNA, don't have to actually accomplish anything as long as they can propagate themselves.
aogaili 6 hours ago||
But they need headcount under them to justify their position...
trumpdong 6 hours ago||
That's not "accomplishing anything"
aogaili 2 hours ago||
Yes, but I didn't mention that they accomplish anything? mid-managers role is not to accomplish is to manage the team that accomplish..duh
himata4113 16 hours ago||
While LLM's are beyond junior level at this point, they're still just that. I don't really agree that the first two pillars have been affected.

I've shared a story before that between now and 2 years ago a developer who solely relied on AI has produced the same hot garbage instancing system within the same time period. For example back in my day in 2 years I went from writing a system that struggled with few hundred players to one that could handle thousands and far beyond that. The person using AI 2 years ago wrote a system that didn't work and wrote a system 3 months ago that doesn't work.

Everyone is saying how great AI is, but they're missing that the driver is just as important AI wouldn't be able to achieve any of this without capable (often seniors) using it and giving it guidance. It's really a difference between "it works" and "it works without flaws".

Of course AI can produce things that also "work without flaws" with solved problems and someone "recreating" something that already exists with AI is not that special, a junior developer could accomplish the same thing given the time.

But I do agree that AI becoming part of performance reviews and all that is producing more productive developers which is going to drive the cost way down. In a way AI is stealing from a developers salary and giving it to the AI companies which is pretty ironic considering how cold developers seem towards artists.

Melatonic 11 hours ago||
Software engineers really need to unionise. ASAP.
internet2000 15 hours ago||
This is good. We want less barrier to entry and more competition in software.
phyzix5761 16 hours ago||
LLMs are good at general solutions but not specific solutions. As industries evolve and laws, regulations, and practices change LLMs will struggle because those things are not included in its training set yet. We'll always need humans to push companies in new directions in order to compete, unless we eradicate capitalism altogether and then we're all out of luck. No competition means no incentive to try and be better than the next guy which means no new products and services for humans to develop that AI hasn't seen already.
hmokiguess 15 hours ago|
One other thing I find it is bound to happen is that this domain knowledge you speak of is just going to shift towards LLM domain knowledge.

Look at prompt engineering, and how quickly it became a hot thing. Does everyone know to steer their AI well? There's only so much a harness can do for you once you start attempting to one shot with a single sentence of 4 words.

As others said, "write a Rust compiler make no mistakes" can only work if you overfit a harness to that single prompt. Nobody is going to do that.

So the part you mentioned about the knowledge you accumulated around how to know that "trade-offs between implementations" and "idempotency to prevent double-charges" is just moving to the domain of the english language and tokenizers. One could argue here that this is far more interesting as it requires you to explore deeper into how we communicate and describe the world around us. Reminds me of physics and math.

I think there's an optimism lenses to it if you can grasp it as an opportunity rather than an inevitable doomsday apocalypse.

More comments...