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

Vibe coding and agentic engineering are getting closer than I'd like

simonwillison.net

Post reply on HN
Print view View original

841–850 of 958 posts

Re: Vibe coding and agentic engineering are getting closer than I'd like

#841

Post by zarkov99 »

sothatsit wrote: post #606
ori_b wrote: post #545

Earlier quoted context omitted.

How much calculus do you think you could pick up skimming a textbook without doing exercises? We mocked these "architects" from experience. We knew that if you weren't feeling the friction yourself, you wouldn't learn enough to do good design. Maybe you don't care about engineering great systems. Most companies don't. It's good for profit. This isn't new, though AI enables less care.

The entire mistake you are making is comparing using AI to skimming textbooks, or taking shortcuts. Your entire premise is wrong. People who care about craft will care about the quality of what they produce whether they use AI or not. The code I ship now is better tested and better thought through now than before I used AI because I can do a lot more. That extra time goes into additional experiments, jumping down mor…

I care deeply about craft, but:

a) I cannot effectively review more than 2000 lines of code a day. The LLMs can produce much more than that. b) Even if I accepted my reading throughput limitations as the cost of being in the loop, reading is not enough to keep cognitive debt in check: my skills will atrophy if I do not participate in the writing ("What I cannot create I cannot understand").

So, to me, it seems like we, humans, either have to come up with higher (and deterministic) abstractions than code to communicate with LLMs or resign ourselves to letting the LLM guess what we want from English and then banging on the output to see if it sort of works. This later state of affairs seems to be what the current trend is and I find that absolutely revolting.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#842

Post by ewild »

MrScruff wrote: post #771
stephenr wrote: post #747

Earlier quoted context omitted.

.... that an IDE providing a suggestion about what comes next as you type is not new, and the entire basis of how an LLM works is "what word probably comes next". I'd have thought someone who's so enamoured with the tech would have at least a basic understanding of how it works.

Indeed. To be honest, I think everyone on HN is aware of how LLMs work at this point, it’s not actually adding a great deal to the discussion to keep going on about autocomplete or ‘stochastic parrots’.

At this point if someone calls it auto complete they can be written off as a Luddite with nothing valuable to say. The irony being they themselves are being a stochastic parrot by parroting the jargon other people say about llms.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#843

Post by r4nd0m4ch »

kyykky wrote: post #538
justech wrote: post #512

Earlier quoted context omitted.

I don’t think it’s going to burst like how other people expect. The technology is already out there, when it loses steam people aren’t suddenly going to stop using it. I predit it’ll be more like the dot come crash where companies that can survive the downturn come out dominant.

It ends like this: all codebases become unmaintainable spaghetti after agentic AI spends years on it. Then after every agent in existence will spend minimum 24 hours reading the codebase to add a simple feature, the software is abandoned.

no way it's going to happen

Re: Vibe coding and agentic engineering are getting closer than I'd like

#844

Post by eli »

Daishiman wrote: post #541
eli wrote: post #514

Earlier quoted context omitted.

A much more charitable framing: people who enjoy the process vs people who enjoy the result. (Though, granted, the results are a lot better if you craft it by hand)

I am not really sure. I wrote some scripts that aggregated data from several APIs with an LLM and the LLM had the foresight to create a caching layer for the API responses as it properly inferred that I would need the results over and over again as well as using asyncio to accelerate fetch speed. This would have been a v2 or v3 and it one-shotted it perfectly.

When it works, it feels genuinely miraculous. Working in a common problem space, like gluing together APIs, it generally does well. Doing something novel or even a little complicated, it can really lead you astray.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#845

Post by mrothroc »

The "blurring" framing makes Simon's tension sound intrinsic when it is actually structural. Vibe coding and agentic engineering aren't on a continuum. They're distinguished by the process.

Engineering is always about a defined process. We follow it to produce predictable artifacts that meet the specifications. Even though code is somewhat "squishy" in that it is an art just as much as a science, it still has to meet the spec.

This has always been true, even before agents started writing code for us. We've all dealt with spaghetti code because of undisciplined practices. That's exactly why we came up with the standard SDLC process: plan, design, code, test, deploy. Repeat.

The part people seem to forget about when looking at this is the space between the steps: the gates. We review the artifacts produced at each stage. If the reviewer does not approve, the engineer has to fix it until it passes. True for human coders, doubly true for agentic coders.

Agentic engineering still follows the process. Artifacts are now cheap to produce, which means we have to adjust it so we don't overwhelm the humans in the loop. For me, this means augmenting my review step with agentic reviewers to catch the dumb stuff. It only escalates to me when either a) it passes clean or b) there is something that genuinely needs my experience.

This is agentic engineering, not vibe coding.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#846

Post by ex-aws-dude »

wg0 wrote: post #365

Here's for the AI supremacists: Let's assume AI is 10x perfect than humnas in accuracy and produces 10x less bugs and increases the speed by 1000x compared to a very capable software engineer. Now imagine this: A car travels at a road that has 10x more bumps but it is traveling 1000x slower pace so even though there are 10x bumps, your ride will feel less bumpy because you're encountering them at far lower pace. Now…

I don't understand what you mean by the last point

If I generate code with an agent and review it and iterate back and forth until the quality is as high as I would write myself, the end result is no different

I'm still in control of holding it to the same quality level?

With agentic coding there is still a human reviewing the code, that's the main difference from vibe-coding

The rules are just to try to guide it and save iteration time but there is no illusion that they are actual hard rules since everything is statistical.

> Such paradigm assumes that LLMs are perfect reliable super accurate rule followers

That's the whole reason we're not vibe-coding, we are well aware of that.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#847

Post by sothatsit »

zarkov99 wrote: post #841
sothatsit wrote: post #606

Earlier quoted context omitted.

The entire mistake you are making is comparing using AI to skimming textbooks, or taking shortcuts. Your entire premise is wrong. People who care about craft will care about the quality of what they produce whether they use AI or not. The code I ship now is better tested and better thought through now than before I used AI because I can do a lot more. That extra time goes into additional experiments, jumping down mor…

I care deeply about craft, but: a) I cannot effectively review more than 2000 lines of code a day. The LLMs can produce much more than that. b) Even if I accepted my reading throughput limitations as the cost of being in the loop, reading is not enough to keep cognitive debt in check: my skills will atrophy if I do not participate in the writing ("What I cannot create I cannot understand"). So, to me, it seems like w…

I think the distinction is that for experiments and prototypes the behaviour of the final system is what we are trying to design. We can experiment and see the tradeoffs and explore the design space before committing to a direction. And then we can sit down and produce the final code to a quality we are happy with. If you are serious about this process, there is no way you are producing 1000s of lines of code a day, unless it is trivial boilerplate.

In terms of higher-level abstractions, I agree this is one particularly treacherous rung on the ladder of abstractions. Previous abstractions like compilers or garbage collectors have at least had more structure/rules to rely upon. I don't know exactly how that will look but I don't think we will solely be relying on banging on the output, we will also be spot-checking the source code, using profilers or other tools to inspect the behaviour of systems, and asking the agent to explain the architectural decisions made. I'm not sure exactly how this will look, but I do believe that people who care will still find ways to do good work.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#848

Post by codebolt »

ori_b wrote: post #820
sothatsit wrote: post #606

Earlier quoted context omitted.

The entire mistake you are making is comparing using AI to skimming textbooks, or taking shortcuts. Your entire premise is wrong. People who care about craft will care about the quality of what they produce whether they use AI or not. The code I ship now is better tested and better thought through now than before I used AI because I can do a lot more. That extra time goes into additional experiments, jumping down mor…

You can keep telling yourself that. I have seen the results from others making the same arguments. The result is invariably trash.

My agentic workflow probably differs somewhat from the majority of others here, but I can positively guarantee you that both the quality and quantity of my output is significantly higher than it has ever been, in my 20-something years of writing code. And at least 90÷ of the code I've written this year was output by an LLM. You can keep sticking your head in the sand, in the end it will only be to your own detriment.

Re: Vibe coding and agentic engineering are getting closer than I'd like

#849

Post by redsocksfan45 »

sarchertech wrote: post #500
eddieroger wrote: post #416

Earlier quoted context omitted.

> If I know the answer to all these questions, wiring it together takes me LESS time than passing it to Claude Code. That's just not true, and if it is in your case, then you're not great at writing prompts yet. > Take the todo_items table in Postgres and build a Micronaut API based around it. The base URL should be /v1/todo_items. You can connect to Postgres with pguser:pgpass@1.2.3.4 That's about all it takes these…

>you’re not great at writing prompts yet How do you reconcile that with your example prompt, which demonstrates no skill requirement whatsoever. It’s the first thing any developer would think of.

[dead]

Re: Vibe coding and agentic engineering are getting closer than I'd like

#850

Post by ttariq »

I am not sure about agentic engineering getting close to vibe coding, but I certainly buy into building trust in your agents, similar to how you would trust another team / colelague within your organization (the image resizing example), and the best way to make sure that a team is working well is to make sure the right context i available to them at the right time and whenever they change the code base, they update that "context." In the case of human programming, this context is in the form of architecture docs, tickets, product spec, ADRs, messages, code review comments etc and lives in a host of different places. It is also difficult to get humans to fetch and update the context with discipline. However, with agents, it is much easier to get them to consume the right context and keep it updated as they make changes to the code base. I think that is the key to making agents more reliable and being able to have the trust in their decision making and output. All of this, is of course, on top of standard unit testing etc.
Post reply on HN