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

831–840 of 1001 posts

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

#831
post #590

Earlier quoted context omitted.

I have worked on highly regulated areas in finance (risk). Compliance is a highly creative art, often requiring lots of out-of-the-box thinking and non-obvious solutions. The people I found worst at this were IT. They tend to over-interpret regulation, and super-restrict beyond what is needed for actual de-facto compliance. My guess is the model makes the same mistakes as the programmers: taking 'rules' literally, un…

Ha. I've worked in a fairly strongly regulated sector (energy, in the Netherlands), where I collaborated closely with our head of compliance, and she heavily over-interpreted the regulations while I often tried to find more pragmatic solutions. I think adherence to regulation and compliance is nothing to do with whether you're a SWE, a risk officer, or C-level, and everything to do with your own principles, ethics, p…

I've found two things to matter:

1. experience, i.e. knowing why and how a rule matters (in general, but also to auditors)

2. willingness to think

If these aren't present, you get overly restrictive compliance that at the same time accomplishes nothing.

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

#832

Earlier quoted context omitted.

Learning a company and it's product is so natural to us that we hardly talk about it. It's a key skill for reliable workers. It's probably impossible for LLMs to learn and apply that wisdom reliably .

Why? Humans learn this information from documentation and being exposed to all the different systems. We know LLMs can ingest documentation like a sponge and access all different internal systems, resources and products via MCP Servers?

Of course they can, but if you give a LLM a specific rule, what happens is that it only shifts the probability of following the rule. Prohibiting a rule violation is technically impossible via prompting.

Humans do make mistakes or forget things as well. We learn to not rush on stairs and to not touch hotplates. A few bruises later that wisdom is permanent, at some point we don't even need to fail with everything to accept the rules.

A LLM is permanently at risk to break every given rule.

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

#833

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…

Nothing corroborated this. Performance on benchmarks has practically leveled off. The big gains have come from architecture (have a secondary LLM review output) or searching the internet. Also prices are going up. Everything points to the likelihood that we're at the top of the curve.

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

#834
post #392

Earlier quoted context omitted.

The dynamic of agent codes human reviews does seem like the only sane one for the foreseeable future. Even Anthropic themselves still fall back to this. The problem is that sucks, even if all software engineers keep their jobs and salaries, the floor is still pulled out from under us. Imagine if a surgeons job was to supervise robot surgeons from a remote computer, or a woodworker just signs off on work before the ma…

"Soul and humanity" is doing a lot of work here. Does the woodworker who shape using a handsaw use less "soul" than the one who uses a machine? Does the musician who use a DAW and VSTs instead of analogue tape recorders create music with less "soul"? Does the painter who buys acryllic paint instead of synthesizing their own dye from plants use less "soul"? As technological innovation progresses, the barrier to creati…

Hmm, but AI isn't just a really-good tool, it's also doing creative work too. As an aspiring composer, music generators are in my opinion really quite good, and often matches what composers and song writers can do. So if you ask me does a person who creates music with gen AI miss out on the soul of creating music? In my opinion at least, for the most part, yes.

To write a piece of music, you're working at so many different levels, the analytical, emotional, and structural as well as drawing on years of training and experience. When a person with little (or even just a moderate amount of) music training generates a piece of music in a couple hours, are they actually a composer? I personally would say no. I mean it does take a good ear (which is important) to use a music generator well, but still, would say they are more of an editor or an evaluator or a practical critic of music instead of a composer.

Yeah, at what point in a discipline does the increasing skill of AI overtake the need for our contributions from the working being done? In music, it's happened already. Looks like it's happening in coding too.

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

#835

I always remember of the infamous Steve Jobs quote "Ideas are cheap". If execution is everything, and frontier LLMs solve execution, then ideas are the gateway to abundance now, but abundance alone does not guarantee "stickiness". What I think is often overlooked is the human "Willingness" and "Care" of staying with the thing for the lack of a better term. What I mean by that is that a lot of people just don't care e…

LLMs don't solve _all_ of the execution. Having distribution and clout is even more important to execution now. To win people must take you seriously when you ask for business partnerships or to give your product a try. The winners have a lot of clout with plenty of followers willing to try the product or integrate.

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

#836

Earlier quoted context omitted.

Posting this under a burner so I don't dox myself: I work in FinTech on a regulated product. We have access to Mythos. Mythos identified part of our codebase that it confidently asserted was not complaint with a particular regulation and we were at grave risk by allowing it to operate the way it was. Except this was not the case, it had of course hallucinated what the regulation actually required (I know this because…

The dynamic of agent codes human reviews does seem like the only sane one for the foreseeable future. Even Anthropic themselves still fall back to this. The problem is that sucks, even if all software engineers keep their jobs and salaries, the floor is still pulled out from under us. Imagine if a surgeons job was to supervise robot surgeons from a remote computer, or a woodworker just signs off on work before the ma…

Only sane dynamic is human writes code, human reviews code, ai also reviews code.

I don't believe there is any point in having ai generate code

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

#837

Earlier quoted context omitted.

Why would it stop with just developer layoffs? When software companies rely on LLM providers to run their business, I’d argue we‘ll see a massive bust of these companies around the world - from on-prem products to SaaS. Customers may build the software they need entirely in-house or via prompt-engineer consultants, without the need to buy software tools like today. It could be a very very different world.

This won't happen in most cases because the valuable thing is largely the knowledge encoded in the software, which the buyers of the software don't have and don't want to have since they're focused on their own business. There's also, of course, the not insignificant value in the software itself actually working, being operated, being updated when necessary, all of that. Again just extra hassle no business will want…

agreed. Also, data security, data compliance, legal, customer support, operations... Yeah, SaaS is not going anywhere soon.

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

#838
There is no way that LLMs can replace human engineers.

They're often wrong, dumb, and would turn any code into an unmaintainable mess in days if left unchecked.

I use LLMs every day because I can do less work at my job, but they suck and I would never use them for a personal project besides for very isolated and self-contained components.

It's a lot of marketing and hype.

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

#839

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…

[flagged]

Not all work is the same. Some of it is debugging, some is testing, rubber ducking, etc. LLMs reduce effort on such types of work, and the range of what they can handle keeps growing. If you don't see that yet, that's exactly why you should invest the time and mental effort to learn these AI tools deeply.

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

#840

This doesn’t read to me as someone who is sincerely impressed or rather surprised what ai is capable off. This reads like someone is trying to convince me, that ai is just this good, and that the author is telling me to use more ai. To me this sounds like: Trust me, it’s really bad, i know what I’m talking about. Just lean into it, or change profession.

I find this doomsday interpretation of AI replacing engineers ridiculous. Lots of it also seems to be bot-written spam.

Just because an AI seems to know finance/architecture/debugging doesn't devalue authors knowledge of said domain. Its essentially like your coworker having the same knowledge. There's space on the market for both. And if AI gets to the point that it can in fact replace engineers (somehow also claiming accountability?) I'd expect it to be priced competitively enough to make every manager think hard whether to buy and become hopelessly dependent on some 3rd party service or hire an engineer. The market is not a zero sum game.

Post reply on HN