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

Agentic Coding Is a Trap

larsfaye.com

Post reply on HN
Print view View original

231–240 of 403 posts

Re: Agentic Coding Is a Trap

#231

Post by CodeShmode »

ryandrake wrote: post #168
CodeShmode wrote: post #123

Earlier quoted context omitted.

Or you can have a conversation with an agent to build up a requirements/plan spec, asking it to analyse existing code patterns. When it seems like the agent has a good understanding of what needs to be done and how. Ask it to implement, keeping changes as a local spike. Ask the agent questions about all the other teams' code, reaching out to them for questions it can't answer or clarification. With agent capabilities…

I have no idea what a “spike” is, but this seems like a good way to build software when there are no consequences to getting it wrong. What happens if this “spike” violates someone’s patent or puts the company at legal/regulatory risk? What if it leaks users’ personal information? What if it introduces a vulnerability that my 13 year old can exploit? What if it crashes 2 million of your users’ devices because it didn…

The point not to YOLO in vibe coded slop. It's that with agentic generation you can prove out concepts, features and technical approaches first because it's low cost. Once you have something working you can work on getting alignment.

And yes you should completely grok what you're finally committing, making sure it's fully tested. What it enables is lots of fast experimentation before settling on your final change.

https://www.google.com/search?q=software+spike

Re: Agentic Coding Is a Trap

#232

Post by prplxd_nihilist »

ivolimmen wrote: post #222

Slightly off-topic. The funny thing is that he mentions that Spec Driven Development is the future. Technically we allready did this when we where doing Waterfall. I kind of miss it that we had good documentation. The last decade (maybe more?) I get jira tickets with a one liner. I often need to call people as they specify almost nothing. I am still avoiding working with AI. I try to use some models locally for exper…

Upvote. These remote models are going to get ever expensive. Their business future depends on the expected advancements in inference costs. One should never want to tie themselves to this prison.

Re: Agentic Coding Is a Trap

#233

Post by fnordpiglet »

crjohns648 wrote: post #88
fnordpiglet wrote: post #51

Interestingly I’ve learned more about languages and systems and tools I use in the last few years working with agentic coding than I did in 35 years of artisanal programming. I am still vastly superior at making decisions about systems and techniques and approaches than the agentic tools, but they are like a really really well read intern who knows a great deal of detail about errata but have very little experience.…

I have also seen the learning acceleration, there's a significantly increased set of techniques and technologies I have learned how to apply. From a person perspective though, I'm apprehensive about the effect AI will have on the human "very well read intern." People who know a lot very deeply about specific areas are fascinating to talk to, but now almost everyone is able to at least emulate deep knowledge about an…

I agree. I think this is personally very useful, but I think a great deal of what made computing an amazing industry to work in is going to or has already died. I suspect the general field as it existed will entirely cease to exist before the current set of well read interns have had much of a career chance. It is sad, but we have finally succeeded in programming ourselves out of jobs.

Re: Agentic Coding Is a Trap

#234

Post by fnordpiglet »

liamdgray wrote: post #47
fnordpiglet wrote: post #42

Earlier quoted context omitted.

I’ve build a configuration transpiler to Claude code and codex and found I can switch pretty quickly between both and run both at once. At the moment codex performs better. Prior CC did. There is no vendor lockin and this is an old canard in technology that LLMs in fact themselves make irrelevant. Once you’ve got an implementation that uses X converting it to Y is almost trivial with an LLM because the spec is canoni…

Great idea. I would love to see your transpiler Mind sharing it?

It’s buried in my dotfiles and not easily extracted. But the idea isn’t a hard one to implement, except the coding engineers are woefully unaware of themselves. Codex is easier because it’s open source. Claude you kind of have to futz with it for a while. Once you have the intermediate form working and outputting config for the two I’m sure you can coerce it to any other agent that comes along with similar constructs (marketplaces, etc). Theres some nuance for some MCPs particular those that download binaries like rust MCPs but its very complex I found and probably better to avoid unless you really need it.

Re: Agentic Coding Is a Trap

#235

Post by dilyevsky »

nyssos wrote: post #220
dilyevsky wrote: post #132

Earlier quoted context omitted.

Respectfully, unless someone is really really bad at articulating what the quality standards are or works with a very niche stack that is definitely not the case anymore with SOTA models

You're presuming too much about what OP's quality standards are. Can SOTA models outperform the average junior engineer? Yes, obviously. Can they match the best human engineers, if those humans were given all the time and interest in the world? Equally obviously not. I use hundreds of millions of tokens a month, and LLMs have completely transformed the way I work. They're also, frankly, pretty mid programmers.

most people are around mid by definition. mid is good. a lot of companies pay good money to find mid devs.

Re: Agentic Coding Is a Trap

#236

Post by otabdeveloper4 »

dilyevsky wrote: post #213
otabdeveloper4 wrote: post #209

Earlier quoted context omitted.

Because "platform engineering" isn't about writing bash scripts, it's about having a mental model of the system architecture. (Which the LLM definitely doesn't have.) Also, it's always the case where you think LLMs are great at doing whatever it is that you don't understand or value.

There was lot of lofty ideas around devops/pe and the like but ime the grim reality is that in most of managers’ heads it’s basically “tools that product engs don’t want to be concerned with” and like i said current models are really really good at this. Also disagree on the mental model thing - give opus 4.7 or latest codex access to your code, design mds and your logging mcp - in five minutes it will be better at d…

> and like i said current models are really really good at this

They're absolutely shit at this. You only say that because it's a thing you "don't want to be concerned with".

> in five minutes it will be better at debugging production issues

In my circles "debugging production issues" is running perf to diagnose memory allocation hot paths and tcpdump to figure out who is sending bad packets.

Re: Agentic Coding Is a Trap

#237

Post by utopiah »

keyle wrote: post #79

As a senior developer, 25+ years, I have been thrown recently into a meeting "hey can you join in for 5 mins". I really don't like these meetings where you're dragged in in the middle of them without any clue. The questions came flying in fast, without any introduction, and this was about an external integration out of a dozen. They have their own lingo, different from ours, to make the situation worse. I had a _very…

Back to 2014 with this 7min "The Expert" skit https://www.youtube.com/watch?v=BKorP55Aqvg

Not the kind of meetings I enjoy where basically expertise isn't seen as something to build on but rather just (creative) confirmation bias.

Re: Agentic Coding Is a Trap

#238

Post by dilyevsky »

otabdeveloper4 wrote: post #236
dilyevsky wrote: post #213

Earlier quoted context omitted.

There was lot of lofty ideas around devops/pe and the like but ime the grim reality is that in most of managers’ heads it’s basically “tools that product engs don’t want to be concerned with” and like i said current models are really really good at this. Also disagree on the mental model thing - give opus 4.7 or latest codex access to your code, design mds and your logging mcp - in five minutes it will be better at d…

> and like i said current models are really really good at this They're absolutely shit at this. You only say that because it's a thing you "don't want to be concerned with". > in five minutes it will be better at debugging production issues In my circles "debugging production issues" is running perf to diagnose memory allocation hot paths and tcpdump to figure out who is sending bad packets.

> In my circles "debugging production issues" is running perf to diagnose memory allocation hot paths and tcpdump to figure out who is sending bad packets.

Haha i got some news for you…

Re: Agentic Coding Is a Trap

#239

Post by utopiah »

It's funny because it's either

- tech for management who can delegate but don't have the expertise to know when it's wrong or just plainly impossible

- tech for coders who have the expertise... but who will gradually lose it.

So I'm not sure who it is for, beside VCs and shareholders until the next quarter obviously.

Re: Agentic Coding Is a Trap

#240

Post by dilyevsky »

SpicyLemonZest wrote: post #223
dilyevsky wrote: post #218

Earlier quoted context omitted.

I think it’s because most PEs are in general conservative (the whole “mr no” meme) and also have limited experience writing software so they will be slow to adapt to this paradigm shift

It's not my experience that our production engineers are quick to say no or slow to adopt LLM agents.

There’s someone in the sibling thread arguing claude is bad at using perf and tcpdump
Post reply on HN