Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Six months of writing code exclusively with agents

blog.exe.dev

Post reply on HN
Print view View original

91–100 of 110 posts

Re: Six months of writing code exclusively with agents

#91

Post by monospacegames »

redlewel wrote: post #33

How do people read these types of post with this AI flair, I couldn't read more than a couple sentences

It is such an incredibly disrespectful way of communication.

If the author can't bother to write the article himself why should the thousands of people who come across this post bother to read it?

How do I know where the author's own thoughts end and AI's assumptions about his thoughts begin?

Then again people typically don't read past the title so I suppose it doesn't matter what the content of the article is, as long as it gets people talking and directs traffic to their website.

Re: Six months of writing code exclusively with agents

#92

Post by a2ff6eeb0 »

ragequittah wrote: post #72
jplusequalt wrote: post #64

Earlier quoted context omitted.

>I think you will always need a button pusher. Okay. Is that job going to provide a competitive salary, with good benefits? More importantly, is it going to be personally fulfilling? Software developers who have adopted agentic coding are worse than class traitors. They are actively sabotaging the entire careers future viability, and have made me ashamed to think anyone in this industry was worthy of the title "engin…

"Compositors who have adopted Linotype machines are worse than class traitors. They are actively sabotaging the entire trade’s future viability, and have made me ashamed to think anyone in this industry was worthy of the title “printer.”" "Weavers who have agreed to tend a fourth loom are worse than class traitors. They are actively sabotaging the entire trade’s future viability, and have made me ashamed to think any…

Yes. The only rational response to the coming AI wave is to get out of the business of selling intelligence for money. The machines will out-think you for cheaper.

Re: Six months of writing code exclusively with agents

#93

Post by bossyTeacher »

jghn wrote: post #74
beej71 wrote: post #69

Earlier quoted context omitted.

Disclaimer: I'm an older dev out of the game. I could definitely see the bigger picture faster than I could type it (except for challenging tasks), but that typing time was also synergistic thinking time. Not unfrequently, I'd find a better design while writing the code. Even if it didn't make it into that codebase, it could be filed away for future reference. Edit: I recall reading somewhere that John Carmack doesn'…

This is part of what I don't understand about the new narrative. I agree in principle that yes, the mechanical act of writing code was always the real bottleneck. But people act like all you have to do is carefully plan everything out up front, double check it at the end, and bob's your uncle. I always found that the act of chugging the code helped to refine my thoughts along the way, including the big picture of it…

>But people act like all you have to do is carefully plan everything out up front, double check it at the end, and bob's your uncle.

It's a coping mechanism. Goal shifting to protect yourself.

Re: Six months of writing code exclusively with agents

#94

Post by bossyTeacher »

a2ff6eeb0 wrote: post #34
iamflimflam1 wrote: post #18

Writing code with LLMs takes discipline and is difficult. You need to learn how to do it and how to get the best from it. Software engineering is dead, long live software engineering!

It really isn't that hard. The LLM needs a few iterations of self review, and you need to ask it for an architecture up front, but from there, it's mostly manual testing to make sure the LLM didn't fuck up. I don't think it'd take more than a few weeks for a high schooler to learn how to prompt well.

> I don't think it'd take more than a few weeks for a high schooler to learn how to prompt well.

Devs are notoriously bad at estimating tasks.

Re: Six months of writing code exclusively with agents

#95

Post by queenkjuul »

jghn wrote: post #74
beej71 wrote: post #69

Earlier quoted context omitted.

Disclaimer: I'm an older dev out of the game. I could definitely see the bigger picture faster than I could type it (except for challenging tasks), but that typing time was also synergistic thinking time. Not unfrequently, I'd find a better design while writing the code. Even if it didn't make it into that codebase, it could be filed away for future reference. Edit: I recall reading somewhere that John Carmack doesn'…

This is part of what I don't understand about the new narrative. I agree in principle that yes, the mechanical act of writing code was always the real bottleneck. But people act like all you have to do is carefully plan everything out up front, double check it at the end, and bob's your uncle. I always found that the act of chugging the code helped to refine my thoughts along the way, including the big picture of it…

I can't even count how often i think i have everything planned out and then halfway through implementation realizing a much better way, or some case i didn't consider, or a different way to eliminate the problem, etc. This is true both when I'm writing code myself and when I'm driving Claude. I see coworkers just taking the first concept from Claude and running with it and you can really tell that skipping the thinking that goes on during the implementation step really has drastic bad effects on the final code.

Re: Six months of writing code exclusively with agents

#96

Post by bossyTeacher »

alexpotato wrote: post #71
phyzix5761 wrote: post #68

Writing code used to be free, anyone could do it with enough patience and knowledge, but now you have to pay hundreds of dollars to do this kind of thing with agents. It seems the barrier to entry has gotten more expensive.

It's still free to write code... e.g. I can ride a bike for free compared to driving a car where I have to pay for gas. And their are pros and cons to both. e.g. bikes are easier to park but not great in bad weather and vice versa

Not quite.

In a car world, riding a bike is free but acquiring a bike is incredibly expensive because they are not mass produced and are instead a luxury that rich people pay for to get a single bike manufactured for them.

The equivalent is that in the future, most open source dies, and the little open source software that remains is low-level, systems critical. And so, if you want to write code, you must first use an LLM to make you a compiler for your desired language that works in [insert future pc architecture] as any of the old 2026 era compilers stopped working decades ago. If you don't have enough tokens for that, tough luck.

Re: Six months of writing code exclusively with agents

#97

Post by queenkjuul »

keeda wrote: post #75
kelseyfrog wrote: post #16

Earlier quoted context omitted.

I've heard this reported, but it's hard to imagine it. The biggest bottleneck for me has always been hammock time[1]. 1. https://github.com/matthiasn/talk-transcripts/blob/master/Hi...

Yes, but once you have the solution crystalized in your head do you not feel some frustration in not getting it implemented as soon as possible? I often found myself impatient in getting something I knew would work to actually work, and it was always due to the tiresome fractal of complexities that any non-trivial task involves, like looking up an API or working around the quirks of some dependency. On a related note…

> Yes, but once you have the solution crystalized in your head do you not feel some frustration in not getting it implemented as soon as possible?

For me, honestly, no. Once i have the solution i have the solution. The only people that care about how fast it gets published are the project managers at work, certainly not me.

And besides, writing it all out and confirming it all along the way seems helpful to me, it ensures i understood everything correctly and sometimes inspires new ideas as well

Re: Six months of writing code exclusively with agents

#99

Post by a2ff6eeb0 »

bossyTeacher wrote: post #94
a2ff6eeb0 wrote: post #34

Earlier quoted context omitted.

It really isn't that hard. The LLM needs a few iterations of self review, and you need to ask it for an architecture up front, but from there, it's mostly manual testing to make sure the LLM didn't fuck up. I don't think it'd take more than a few weeks for a high schooler to learn how to prompt well.

> I don't think it'd take more than a few weeks for a high schooler to learn how to prompt well. Devs are notoriously bad at estimating tasks.

I have a non-technical cousin that figured it out on their own just fine.

Re: Six months of writing code exclusively with agents

#100

Post by queenkjuul »

a2ff6eeb0 wrote: post #43
iamflimflam1 wrote: post #38

Earlier quoted context omitted.

Maybe - but I would argue that the skill comes in knowing if the architecture is appropriate and knowing that LLM is not screwing it up.

You can get more confidence by asking the LLM about when the architecture would be appropriate than by trying to intuit it yourself. The LLM is trained on so many more architectures and stories of how they panned out than you are. The skill is really in manually testing before handoff.

I still get occasional hallucinations, wrong answers, and entirely unnecessary change proposals from LLMs. An LLM being confident in its own work does not make me confident in it.
Post reply on HN