[scuff3d]

You've left out an absolutely massive swath between "low level systems" and "crud app". There a lot more to the industry then just those two things.

And as productivity rises, the complexity and ambition of the projects tackeled will continue to rise.

The way we do the work might change, but things aren't going to look as different as people think.

[aadv1k]

Yeah, I think that's a fair point. I mean, we have SRE, which is a whole set of skillsets in itself and it does involve around maintaining "CRUD-apps"; Heck there is the cybersec industry as well, ensuring CRUD-apps aren't exploited. My para was certainly a bit more 2D, I was speaking more from what I might benefit learning now, to get a job, say end of 2026. I have observed many roles like SRE etc, require certain experience which is hard to get by working on your own projects without much traction; similiar case for cybersec but to a lesser extent.

It's this paradox where in order to become a senior engineer, you must get hired as a junior engineer, to learn and observe how production software works, but that is pretty hard these days.

[Imustaskforhelp]

> It's this paradox where in order to become a senior engineer, you must get hired as a junior engineer, to learn and observe how production software works, but that is pretty hard these days.

Yes that's true but I would argue this isn't because of AI itself. Sure it might have accelerated it but I have heard that in the 2020's a lot of people got a lot of jobs in the industry and then its that the market did feel saturated and in a sense, a lot of jobs feel this way in the economy where if someone has it, they continue having it but its becoming harder for new people to enter, there are financials reasons for this too mostly if i remember, interests rates are one of the core reasons

That being said, Even though I have used LLM's a lot, I do not agree with your opinion and I am even younger than you most likely (17)

A lot of people my age/people just going into college would use AI a lot to cheat but I will genuinely try to take it slow to learn things. I will try to use it as a learning tool and not as a crutch. Currently I am still in high school and I get ideas which I wish to implement but dont have the time because of exams so I test them out with LLM's but even I find the whole process frustrating at times and its just, maybe its me but there is a good ceiling that I can process. Some basic crud processes can definitely be in that but if your project is novel, even basic things can be hard

I will give you an example, I recently tried creating an api for proton docs which worked by having browser instances. I had two basic scripts that worked on one platform and converted them to puppeteer using LLM and after it worked, I then asked it to simply create a very basic crud api on top of it

Nope, I tried it 5-6 times on the best models on the market but they couldn't actively take two files which had read/write and have it work, either there was an issue in read or an issue in write

I am fine with the project as it is right now and I am fine with the templates it has given me and I am going to build it now on top of it

I had many ideas which could be considered basic crud apps like having a kanban app/github issues like ui on top of bitwarden after I saw a post here by simon saying how he uses github issues and that went semi-viral on hackernews

Another issue is that I think just as how sure writers and similar can generate AI generated, I feel like there would be more trust on the non AI generated code.

Personally I will try my best to create prototypes with AI and if I like them, I will see what it does and then rewrite them myself as an learning experience and also because people say spec driven development etc. but I just want to code things by hand at this point or convert the LLM generated prototypes of idea to something that I later code and understand by hand tbh (most likely when I get into college)