Live data from Hacker News

LLMs are eroding my software engineering career and I don't know what to do

human-in-the-loop.bearblog.dev

811–820 of 1001 posts

Re: LLMs are eroding my software engineering career and I don't know what to do

#811

Earlier quoted context omitted.

> The first one-shot app was created with ChatGPT in June 2023 - 3 years ago. In my experience, the current result of one-shotting apps is just as bad today as it was back then. Hard disagree. Take a good SaaS starter template and do a bunch of harness engineering. You can get an agent to shit out production grade stuff. You might argue that's cheating, but there's nothing stopping you from doing it, and it works. It…

The original comment asked for someone to name one and you didn’t, though.

Cos I'm too busy doing it at work.

Re: LLMs are eroding my software engineering career and I don't know what to do

#812

I wonder how do people use LLMs so it does not hallucinates. Like 90% of the time the code is impeccable, but the remaining 10%... Let's say I determine the expertise by how well do people act of these 10%. For me, the first pillar is still there, but not in a good condition

Just pretend you are working with a buggy api with poor misleading documentation. LOL Check your assumptions.

Re: LLMs are eroding my software engineering career and I don't know what to do

#813
post #255

There are lots of positives that have resulted from using AI in software engineering. (1) No more long repetitive text editing sessions. I.e. changing namespaces or replacing deprecated APIs with the "correct" ones. AI will make nearly perfect text modifications with ease. (2) No more bike-shedding code reviewers nitpicking over every tiny coding decision. I.e, "you should use std::format instead of std::stringstream…

> Now average folk can solve problems and avoid dealing with Microsoft for a spreadsheet program.

You seriously believe some vibecoder can write anything remotely similar to MS Excel, which took probably thousands man-hours to create?

Re: LLMs are eroding my software engineering career and I don't know what to do

#814

My career path is suprisingly similar to the author's. Weirdly enough, what he takes as the first pillar to fall is the one I see most undamaged currently. LLMs routinely fail at our business specifics: Local tax regulations, particularities of the accounting process, specifics of our ledger implementations. They're great at refactoring, translating between languages, tracing bugs on existing code even, but there is…

> LLMs routinely fail at our business specifics: Local tax regulations, particularities of the accounting process, specifics of our ledger implementations.

So there is a spectrum here, and i dont know what i dont know - meaning i can just be wrong. But we're both on that spectrum and are you sure its not a skill issue?

All of the specifics you list seems so fundamental that in similar projects I've inserted them straight into the AGENTS.md or a strong reference and where to look them up.

If you boil it down to it, you're quite literally saying the problem is the LLMs dont have access to a bunch of facts.

Re: LLMs are eroding my software engineering career and I don't know what to do

#815
I don't know, I've been interviewing a lot lately, for jobs that are specifically looking for engineers keen with AI. But they still drill me on the technical stuff, whether I really have the years of experience I claim.

I guess they don't want just any script kiddie. And that seems very logical to me. So I don't really see the big issue. They still want experts. You just shouldn't do the same stuff you did last 10 years.

Re: LLMs are eroding my software engineering career and I don't know what to do

#816

I see many comments saying, "AI can't do X with 80-100% accuracy; therefore our professions are in good hands." While I don't want to sound overly pessimistic, the models are improving at a rapid rate. If asked ~3 years ago where the state of the models are today, it would sound like sci-fi if answered, "the models are creating full MVP apps in ~30 minutes with one prompt". The hurdles the models are facing now, like…

> the models are improving at a rapid rate. If asked ~3 years ago where the state of the models are today, it would sound like sci-fi Absolutely true, many things will continue to improve in significant ways. However, if we look at the modern history of rapid disruptions driven by technology (a side interest of mine), persistent patterns emerge. Similar to avalanches or flash floods, such periods of very rapid disrup…

> The rate of fundamental, broad-based breakthroughs lifting all LLM applications has clearly slowed with many of the most impactful recent discoveries being in scaling, optimization, tuning and productization toward specific domains.

To me it definitely feels like it's still accelerating, with the most impactful recent discovery being RL training reasoning models (late '24, early '25).

There's an interesting article called "sigmoids won't save you" https://www.astralcodexten.com/p/the-sigmoids-wont-save-you which argues that (unless you have privileged information) you should always assume a process will continue about as long as it’s continued already. (Lindy's Law)

With that in mind the current disruption should last another 10-15 years (assuming it started in '10 or '17.)

Re: LLMs are eroding my software engineering career and I don't know what to do

#817

Earlier quoted context omitted.

> In my experience, the current result of one-shotting apps is just as bad today as it was [three years ago] Hard to respond to this with anything other than "no, you're wrong".

What’s the production mvp app that ai made in 30 minutes then? That was the original question no?

When did you last try this? Here's a prompt for you:

> It's really important to me you make sensible decisions here, and don't bother me with the small stuff. I want a plant-watering app me and my wife can share, that shows who watered which plants in our house. I'll deploy this on my home server with Coolify. The app should be attractive, work both on desktop and mobile. We have a bunch of cases where we have multiples of one plant type. We'll need separate users, but don't go overboard with auth. I want to impress her, so let's lean on the side of more rather than fewer, features, but I don't really wanna run anything that won't just fit in a single container with some persistent storage. We're the only two users who'll ever actually give this a go. Visually attractive is important to me.

Re: LLMs are eroding my software engineering career and I don't know what to do

#818
post #790
post #698

Earlier quoted context omitted.

Those models simply don't hallucinate if you use them properly in any form. The only way they _might_ hallucinate is if you use the web based chat interface and give them zero context.

Ah the old “you’re holding it wrong” argument.

Sorry, but we aren’t at the “press a button to build a game/app” yet.

Re: LLMs are eroding my software engineering career and I don't know what to do

#819
post #34

I sympathise with the author being in the same boat, largely. I just want to emphasise a point... Calculators give 100% correct answers and yet we still hire accountants; for the simple fact that we don't want all to be accountants. People will hire software engineers for the simple fact that they do not want to be software engineers.

funny i was able to do all my taxes this year with ai help and not needing accountant.

That's fine if you have a relatively straightforward situation. If the situation is a bit more complex, with important decisions you should consider, it might still need some eyeballing and time. That is accounting.

Re: LLMs are eroding my software engineering career and I don't know what to do

#820

Wut? I pilot LLMs all day but there's no way in hell I'd agree to be at the helm of a finance product. That first pillar is still there. Maybe the author isn't aware of the impact they have, but I know, with the evidence of reverted PRs, that when I step outside my area of deep knowledge I can no longer call BS on the agents. Our most capable agent, with access to the same kind of distributed systems the author talks…

it doesn't much matter what you and I think about LLM's quality and output

it only matters what upper management think, and its clear that more and more companies value "good enough" and reducing costs over "good"

Post reply on HN