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

Human coders are still better than LLMs

antirez.com

Post reply on HN
Print view View original

541–550 of 762 posts

Re: Human coders are still better than LLMs

#541

Post by realusername »

JeremyNT wrote: post #289
realusername wrote: post #256

Earlier quoted context omitted.

Personally I used them for a while and then just stopped using them because actually no, unfortunately those assistants don't work. They appear to work at first glance but there's so much babysitting needed that it's just not worth it. This "vibe coding" seems just another way to say that people spend more time refining the output of these tools over and over again that what they would normally code.

I'm in this camp... today. But there's going to be an inflection point - soon - as things continue to improve. The industry is going to change rapidly. Now is the time to either get ready for that - by being ahead of the curve, at least by being familiar with the tooling - or switch careers and cede your job to somebody who will play ball. I don't like any of this, but I see it as inevitable.

Maybe there's going to be an inflection point ... or maybe not, I feel like we're at the exact same point as the first release of Cursor in 2023.

Re: Human coders are still better than LLMs

#542

Post by 1718627440 »

JoshCole wrote: post #230
1718627440 wrote: post #228

Earlier quoted context omitted.

Isn't the claim, that it will never make up "C"?

They don't claim that. They say LLMs only generate text someone has written. Another way you could refute their premise was by showing the existence of AI-created programs for which someone isn't a valid description of the writer (e.g., from evolutionary algorithms) then training a network on that data such that it can output it. It is just as trivial a way to prove that the premise is false. Your claim here is sligh…

But doesn't reward for "**C" means that "C" is in the training data?

I am not sure if that is an accurate model, but if you think of it as a vectorspace, sure you can generate a lot of vectors from some set of basevectors, but you can never generate a new basevector from others, since they are linearly independent, so there are a bunch of new vectors you can never generate.

Re: Human coders are still better than LLMs

#543

Post by kweingar »

abletonlive wrote: post #501
kweingar wrote: post #498

Earlier quoted context omitted.

This is why devs who started with J2ME are the holy grail of app developers, since they started making apps years before iPhone devs

you sound mad you could be spending this time upskilling instead. but i'll say it again, when the meta changes the people that were at the top will quickly find themselves at the top again. listen, the reason why they were in the top in the first place and you aren't is a mindset thing. the top are the curious that are experimenting and refining, sharing with each other techniques developed over time. the complacent…

(I deleted a less productive comment.)

I do use these tools though! I spent some time with AI. I have coworkers who are more heads-down working on their projects and not tinkering with agents, and they're doing fine. I have coworkers who are on the absolute bleeding edge of AI tools, and they're doing fine. When the tooling matures and the churn lessens and the temperature of the discourse is lowered, I'm confident that we will all be doing great things. I just think that the "anybody not using and optimizing Codex or Claude Code today is not gonna make it" attitude is misguided. I could probably wring out some more utility from these tools if I spent more time with them, but I'd rather spend most of my professional development time working on subject matter expertise. I want to deeply understand my domain, and I trust that AI use will (mostly) become relatively easier to pick up and less of a differentiator as time goes on

Re: Human coders are still better than LLMs

#545

Post by anonzzzies »

auggierose wrote: post #531
anonzzzies wrote: post #515

Earlier quoted context omitted.

There are so many examples where all current top models just will loop forever even if you instruct them literally the code. We know many of them, but for instance in a tailwind react project with some degree of complexity (nested components), if you ask for something to scroll in it's space, it will never figure out min-h-0 even if you tell it. It will just loop forever rewriting the code adding and removing things,…

This is probably not so much an indictment of the AI, as of that garbage called Tailwind. As somebody here said before, garbage in, garbage out.

Yeah, guess so, but we like garbage these days in the industry; nextjs, prisma, npm, react, ts, js, tailwind, babel, the list of inefficient and badly written shite goes on and on; as a commercial person it's impossible to avoid that though as shadcn is the only thing 'the youth' makes apps with now.

Re: Human coders are still better than LLMs

#546

Post by kweingar »

ukuina wrote: post #537
abletonlive wrote: post #501

Earlier quoted context omitted.

you sound mad you could be spending this time upskilling instead. but i'll say it again, when the meta changes the people that were at the top will quickly find themselves at the top again. listen, the reason why they were in the top in the first place and you aren't is a mindset thing. the top are the curious that are experimenting and refining, sharing with each other techniques developed over time. the complacent…

> when the meta changes the people that were at the top will quickly find themselves at the top again. I think parent is agreeing with you? > This is why devs who started with J2ME are the holy grail of app developers, since they started making apps years before iPhone devs

I was being sarcastic there, a bad habit of mine. There are some advantages to being an early adopter (you get to reap some of the benefits now), but it doesn't give you a permanent advantage, and the people who aren't closely following and adopting weeks-old tools aren't doomed to irrelevance.

The iPhone was an equalizer. Existing mobile devs did get a genuine head start on mobile app design, but their advantage was fleeting.

Re: Human coders are still better than LLMs

#547

Post by wvenable »

codr7 wrote: post #492
wvenable wrote: post #461

Earlier quoted context omitted.

I almost never bother using Google anymore. When I search for something, I'm usually looking for an answer to question. Now I can just ask the question and get the answer without all the other stuff. I will often ask the LLM to give me web pages to look at it when I want to do further reading. As LLMs get better, I can't see myself going back to Google as it is or even as it was.

You get an answer. If that's the answer, or even the best answer, is impossible to tell without doing the research you're trying to avoid.

If I do research, I get an answer. If that's the answer, or even the best answer, it's impossible to tell. When do I stop looking for the best answer?

If ChatGPT needs to, it will actually do the search for me and then collate the results.

Re: Human coders are still better than LLMs

#548

Post by helloplanets »

karn97 wrote: post #369
zachlatta wrote: post #359

Earlier quoted context omitted.

I’d encourage you to try the 1M context window on Gemini 2.5 Pro. It’s pretty remarkable. I paste in the entire codebase for my small ETL project (100k tokens) and it’s pretty good. Not perfect, still a long ways to go, but a sign of the times to come.

[flagged]

> Anytime more than a class or two is involved or if the code base is more than 20 or 30 files, then even the best LLMs start to stray and lose focus. They can't seem to keep a train of thought which leads to churning way too much code.

At least if looking at this specific portion: You can have a 20-30 file code base, with 5-10 classes in the context, in full, and the rest of them filtered in a sensible way. Even a model with a 200k context window can handle this.

The output definitely can stray, but it's not the norm in my experience. Of course, if the output does start to stray, it needs to be snipped in the bud. And the fixes can range anywhere from working but bad code, to very close how you'd written it yourself, if you've clearly described how you want the code to be written.

If you're trying to fix a specific bug, for example, but don't provide thorough logs on what is happening in the code, it's much more likely the output will stray towards some average of what the problem could be, rather than what it actually is in the current code.

And this is absolutely not to say that LLMs could do what Antirez is doing. There is a massive amount of variation in how deeply people think about the code they're reading or writing.

Re: Human coders are still better than LLMs

#549

Post by troupo »

therealpygon wrote: post #399
Buttons840 wrote: post #379

Earlier quoted context omitted.

LLMs aren't my rubber duck, they're my wrong answer. You know that saying that the best way to get an answer online is to post a wrong answer? That's what LLMs do for me. I ask the LLM to do something simple but tedious, and then it does it spectacularly wrong, then I get pissed off enough that I have the rage-induced energy to do it myself.

LLMs follow instructions. Garbage in = garbage out generally. When attention is managed and a problem is well defined and necessary materials are available to it, they can perform rather well. On the other hand, I find a lot of the loosely-goosey vibe coding approach to be useless and gives a lot of false impressions about how useful LLMs can be, both too positive and too negative.

> LLMs follow instructions.

They don't

> Garbage in = garbage out generally.

Generally, this statement is false

> When attention is managed and a problem is well defined and necessary materials are available to it, they can perform rather well.

Keyword: can.

They can also not perform really well despite all the management and materials.

They can also work really well with loosey-goosey approach.

The reason is that they are non-deterministic systems whose performance is affected more by compute availability than by your unscientific random attempts at reverse engineering their behavior https://dmitriid.com/prompting-llms-is-not-engineering

Re: Human coders are still better than LLMs

#550

Post by energy123 »

klabb3 wrote: post #434
retetr wrote: post #424

Earlier quoted context omitted.

Unrelated, but is this a case of the Pareto Principle? (Admittedly the first time I'm hearing of it) Wherein 80% of the effect is caused by 20% of the input. Or is this more a case of diminishing returns? Where the initial results were incredible, but each succeeding iteration seems to be more disappointing?

Pareto is about diminishing returns. > but each succeeding iteration seems to be more disappointing This is because the scaling hypothesis (more data and more compute = gains) is plateauing, because all text data is used and compute is reaching diminishing returns for some reason I’m not smart enough to say why, but it is. So now we're seeing incremental core model advancements, variations and tuning in pre- and post…

We're seeing diminishing returns in benchmark space, which is partly an artefact of construction, not an absolutely true commentary on how things are progressing.
Post reply on HN