Ask HN: In 2024, is SWE a sustainable career? • Hacker News Board

Re: Ask HN: In 2024, is SWE a sustainable career?

#21

Post by karmajunkie » Mon, Oct 07, 2024, 7:03 PM UTC

ilkhan4 wrote: post #6

codingwagie wrote: post #4

My personal opinion is most software will be easy to write by a junior programmer. There will still be high paid roles, but this will be a low paid profession. Web development, UI, etc will be heavily automated with LLMs. You will guide the model in building the software. It's already this way using cursor.sh, most people just havent realized it yet.

I think it will be the opposite, actually. AI seems to be eating the bottom, but there is still demand for higher level decision making that seniors or architects do. The stuff that I would usually delegate to junior devs, e.g. "move this button" or "change this column" are the things it can easily do now and in the near term. But questions like "How should I design this system" are not and I don't see that happening…

this is exactly right. the problem is going to be getting people from junior to senior.

Re: Ask HN: In 2024, is SWE a sustainable career?

#22

Post by karmajunkie » Mon, Oct 07, 2024, 7:03 PM UTC

np- wrote: post #11

ilkhan4 wrote: post #6

Earlier quoted context omitted.

I think it will be the opposite, actually. AI seems to be eating the bottom, but there is still demand for higher level decision making that seniors or architects do. The stuff that I would usually delegate to junior devs, e.g. "move this button" or "change this column" are the things it can easily do now and in the near term. But questions like "How should I design this system" are not and I don't see that happening…

If juniors are being cut out and seniors are being increasingly relied upon to do more and more, that seems to be unsustainable in the long term

welcome to capitalism…

Re: Ask HN: In 2024, is SWE a sustainable career?

#23

Post by danenania » Mon, Oct 07, 2024, 7:03 PM UTC

3D39739091 wrote: post #17

danenania wrote: post #14

Earlier quoted context omitted.

The flip side of this is that AI tools can enable much faster learning, so the time to go from junior to senior can potentially become much shorter for people who are motivated.

If anything, AI tools are doing the opposite. You don't acquire those skills by outsourcing your thinking to an LLM.

It depends how you use them. If your goal is learning and understanding, you will use LLMs much differently than if your goal is to produce copy-pastable code with as little effort as possible.

Honestly, I do a bit of both depending on the situation. Sometimes I just want a one-off script that does xyz and I don't really care enough to have a deep understanding of every detail. Other times I'm getting a crash course in some advanced concept and in an hour I have knowledge that previously only someone with years of experience in a niche domain (and a lot of patience for my dumb questions) could have given me.

Re: Ask HN: In 2024, is SWE a sustainable career?

#24

Post by solarmist » Mon, Oct 07, 2024, 7:04 PM UTC

danenania wrote: post #14

np- wrote: post #11

Earlier quoted context omitted.

If juniors are being cut out and seniors are being increasingly relied upon to do more and more, that seems to be unsustainable in the long term

The flip side of this is that AI tools can enable much faster learning, so the time to go from junior to senior can potentially become much shorter for people who are motivated.

It could (and ultimately I agree it's probably what they should), but that's not what people are seeing juniors do. They're seeing them blindly trust the AI tools and think they don't have to learn what they're doing. So they're building at best a fragle understanding of the code their writing.

Re: Ask HN: In 2024, is SWE a sustainable career?

#25

Post by aristofun » Mon, Oct 07, 2024, 7:07 PM UTC

Is any career a sustainable one anymore?

What makes a career sustainable?

Re: Ask HN: In 2024, is SWE a sustainable career?

#26

Post by danenania » Mon, Oct 07, 2024, 7:13 PM UTC

solarmist wrote: post #24

danenania wrote: post #14

Earlier quoted context omitted.

The flip side of this is that AI tools can enable much faster learning, so the time to go from junior to senior can potentially become much shorter for people who are motivated.

It could (and ultimately I agree it's probably what they should), but that's not what people are seeing juniors do. They're seeing them blindly trust the AI tools and think they don't have to learn what they're doing. So they're building at best a fragle understanding of the code their writing.

Since the results of that approach are predictably bad, I think it will sort itself out. A junior who works like that is already less productive than an LLM in the hands of a skilled engineer, so there isn't really any reason to keep that junior around. But a junior who is using LLMs to rapidly improve their skills can be seen as a good investment even if it takes them some time to become a net positive on the team.

Re: Ask HN: In 2024, is SWE a sustainable career?

#27

Post by dmitrygr » Mon, Oct 07, 2024, 7:14 PM UTC

I am as worried about LLMs taking my job as I am about parrots or magic taking it, and for the same reason. Do not worry about make-believe stuff.

Interviews are irrelevant if you publish enough good things that companies reach out to you and offer to skip them. Write. Learn. Publish. Thrive.

Building fluff is indeed a serious issue, but there are plenty of places that are doing useful things, eg: Medical. Aerospace.

Demand is weak for javascript monkeys. It is plenty strong in the same places it always has been: solving complex problems in complex systems. Go work there.

Re: Ask HN: In 2024, is SWE a sustainable career?

#28

Post by p1esk » Mon, Oct 07, 2024, 7:29 PM UTC

ilaksh wrote: post #18

I've been programming since approximately age 8, so around 40 years. For the past few years have been focused on generative AI, agents, etc. I have built multiple agent frameworks in Node.js, Rust and now Python. I use AI with my agent system to help me write code for projects and my agent system. The leading edge models are incredible, but still somewhat brittle. They can make some weird mistakes. And I definitely h…

The thing that people strangely still seem to not get is that these systems will continue to improve rapidly.

This baffles me as well. So many low hanging fruits to pick, we literally just started with these large models. One obvious direction is predicting the next video frame. Infinite training data is available.

Re: Ask HN: In 2024, is SWE a sustainable career?

#29

Post by aster0id » Mon, Oct 07, 2024, 7:34 PM UTC

ilaksh wrote: post #18

I've been programming since approximately age 8, so around 40 years. For the past few years have been focused on generative AI, agents, etc. I have built multiple agent frameworks in Node.js, Rust and now Python. I use AI with my agent system to help me write code for projects and my agent system. The leading edge models are incredible, but still somewhat brittle. They can make some weird mistakes. And I definitely h…

If AI agents do get better than human SWEs, they'll also get better than their human counterparts at other jobs - including being an entrepreneur.

When people call out the doom of SWE careers at the hands of AI, they seem to think that only SWE careers will be affected, which seems weirdly biased to me. If AI comes for our jobs, it'll come for everyone's jobs. And if that is the case, economics/capitalism as it exists today doesn't make any sense anymore.

Re: Ask HN: In 2024, is SWE a sustainable career?

#30

Post by marssaxman » Mon, Oct 07, 2024, 7:37 PM UTC

I am not worried about it. Software development is the practice of automating away repetitive grunt-work, in every industry possible, especially our own; we have always been engineering our own jobs out of existence, but decade after decade, there continues to be ever more work for us to do.

What has surprised me about the near-apocalyptic enthusiasm for LLMs among some software engineers is that there appear to be a great many people who joined the industry expecting to do repetitive grunt-work, as a matter of course, apparently believing that typing out code is the point of it all and not just a means of expression. I can understand why they might be worried about the future, but perhaps they can relieve their anxiety by developing a deeper understanding of the art.