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

481–490 of 762 posts

Re: Human coders are still better than LLMs

#481

Post by lanstin »

SchemaLoad wrote: post #361
XorNot wrote: post #326

Earlier quoted context omitted.

This was what promptly led me to turning off Jetbrains AI assistant: the multiline completion was incredibly distracting to my chain of thought, particularly when it would suggest things that looked right but weren't. Stopping and parsing the suggestion to realize if it was right or wrong would completely kill my flow.

The inline suggestions feel like that annoying person who always interrupts you with what they think you were going to finish with but rarely ever gets it right.

I'm sorry, it's because of eagerness and enjoying the train of your thought/speech.

Re: Human coders are still better than LLMs

#482

Post by declan_roberts »

Jolter wrote: post #132
bgwalter wrote: post #114

Earlier quoted context omitted.

Yet you are working on your own replacement, while your colleagues are taking the prudent approach.

Here’s the deal: if you won’t write your replacement, a competitor will do it and outprice your employer. Either way you’re out of a job. May be more prudent to adapt to the new tools and master them rather than be left behind? Do you want to be a jobless weaver, or an engineer building mechanical looms for a higher pay than the weaver got?

I would absolutely love to write my own placement. When I can have AI do my job while I go to the beach you better believe I will be at the beach.

Re: Human coders are still better than LLMs

#483

Post by lanstin »

sho_hn wrote: post #285
marcosdumay wrote: post #127

Earlier quoted context omitted.

Most of the times, people are led astray by following average advice on exceptional circumstances. People with a minimum amount of expertise stop asking for advice for average circumstances very quickly.

This is right on the money. I use LLMs when I am reasonably confident the problem I am asking it is well-represented in the training data set and well within its capabilities (this has increased over time). This means I use it as a typing accelerator when I already know what I want most of the time, not for advice . As an exploratory tool sometimes, when I am sure others have solved a problem frequently, to have it r…

I mostly program in Python and Go, either services, API coordination (e.g. re-encrypt all the objects in an S3 bucket) or data analysis. But now I keep making little MPEGs or web sites without having to put in all that crap boiler plate from Javascript. My stuff outputs JSON files or CVS files and then I ask the LLM "Given a CVS file with this structure, please make a web site in python that makes a spread-sheet type UI with each column being sortable and a link to the raw data" and it just works.

Re: Human coders are still better than LLMs

#484

Post by kweingar »

abletonlive wrote: post #463
kweingar wrote: post #287

Earlier quoted context omitted.

If progress continues at the rate that AI boosters expect, then soon you won't have to use them smartly to get value (all existing workflows will churn and be replaced by newer, smarter workflows within months), and everybody who is behind will immediately catch up the moment they start to use the tool.

But if it doesn't and you're not using it now then you're gonna be behind and part of the group getting laid off the people that are good at using these tools now will be better at it later too. you might have closed the gap quite a bit but you will still be behind using LLMs are they are now requires a certain type of mindset that takes practice to maintain and sharpen. It's just like a competitive game. The more in…

If the "meta" changes so quickly, then that sets an upper bound as to how far behind you are, no? Unless you are doing low-autonomy, non-specialized work or are applying to fly-by-the-seat-of-your-pants startup jobs, no hiring manager is going to care if you have three months less experience with Codex than the other candidate.[1]

> so much of using LLM right now is context management

That is because the tooling is incredibly immature. Even if raw LLM capabilities end up plateauing, new and more effective tools are going to proliferate. You won't have to obsess over managing context, just like we don't have to do 2023-level tricks like "you are an expert" or "please explain your thought process" anymore. All of the context management tricks will be obsolete very soon... because AI tooling companies are extremely incentivized to solve it.

I find it implausible that the tech is in a state where full-time prompters are gaining a durable advantage over everyone else. J2ME devs probably thought they were building a snowballing advantage over devs who dismissed mobile development. Then the iPhone came out and totally reset the playing field.

[1] Most employers don't distinguish between three months and nine months of experience with JS framework du jour, no matter what it says on the job listing

Edited to add: Claude Code brought the agentic coding trend to the mainstream. It came out three months ago. You talk about how much you're laughing at the naivete of people here, but are you telling me with a straight face that three months is enough to put a talented engineer "behind"? At risk of being unemployable? The engineers who spent the last three months ping-ponging between Claude Code, Cursor, Codex, etc. can have their experience distilled into like a week of explaining to a newcomer, and I predict that will be true six months from now, or a year from now.

Re: Human coders are still better than LLMs

#485

Post by ben-schaaf »

abootstrapper wrote: post #468
ben-schaaf wrote: post #375

Earlier quoted context omitted.

Friendly reminder that people like you were saying the exact same thing about metaverse, VR, web3, crypto, etc.

I didn’t buy the hype of any of those things, but I believe AI is a going to change everything much like the introduction of the internet. People are dismissing AI because its code is not bug free, completely dismissing the fact that it generates PRs in minutes from a poorly written text prompt. As if that’s not impressive. In fact if you put a human engineer on the receiving end of the same prompt with the same cont…

Again, there were people just as sure about crypto as you are now about AI. They dismissed criticism because they thought the technology was impressive and revolutionary. That it was science fiction come true and only going to continue to improve. It's the exact same hype-driven rhetoric.

If you want to convince skeptics talk about examples, vibe code a successful business, show off your success with using AI. Telling people it's the future and if you disagree you have your head in the sand, is wholly unconvincing.

Re: Human coders are still better than LLMs

#486

Post by emseetech »

roywiggins wrote: post #459
emseetech wrote: post #266

Earlier quoted context omitted.

It's true that manually constraining an LLM with contextual data increases their performance on that data (and reduces performance elsewhere), but that conflicts with the promise of AI as an everything machine. We were promised an everything machine but if we have to not only provide it the proper context, but already know what constitutes the proper context , then it is not in any way an everything machine. Which me…

Depends on how good they get at realizing they need more context and tool use to look it up for you.

How would they reliably recognize the context needed without the necessary context?

Re: Human coders are still better than LLMs

#487

Post by e3bc54b2 »

daveguy wrote: post #354
lupire wrote: post #330

Earlier quoted context omitted.

Which LLMs and which versions?

All. Of. Them. It's quite literally what they do because they are optimistic text generators. Not correct or accurate text generators.

This really grinds my gears. The technology is inherently faulty, but the relentless optimism of its future subtly hiding that by making it the user's mistake instead.

Oh you got a wrong answer? Did you try the new OpenAI v999? Did you prompt it correctly? Its definitely not the model, because it worked for me once last night..

Re: Human coders are still better than LLMs

#488

Post by codr7 »

I would say you thought about this solution because you are creative, something a computer will never be no matter how much data you throw at it.

How did it help, really? By telling you your idea was no good?

A less confident person might have given up because of the feedback.

I just can't understand why people are so excited about having an algorithm guessing for them. Is it the thrill when it finally gets something right?

Re: Human coders are still better than LLMs

#489

Post by bayesianbot »

linotype wrote: post #423
Buttons840 wrote: post #388

Earlier quoted context omitted.

I'm probably suffering undiagnosed ADHD, and will get stuck and spend minutes picking a function name and then writing a docstring. LLMs do help with this even if they get the code wrong, because I usually won't bother to fix their variables names or docstring unless needed. LLMs can reliably solve the problem of a blank-page.

This. I have ADHD and starting is the hardest part for me. With an LLM it gets me from 0 to 20% (or more) and I can nail it for the rest. It’s way less stressful for me to start now.

Been suffering the same, I'm used to having so many days (weeks/months) when I just don't get that much done. With LLMs I can take these days and hack around / watch videos / play games while the LLM is working on background and just check the work. Best part is it often leads to some problematic situation that gets me involved and often I'll end up getting a real day of work out of it after I get started.

Re: Human coders are still better than LLMs

#490

Post by throwaway314155 »

Sohcahtoa82 wrote: post #351
throwaway314155 wrote: post #318

Earlier quoted context omitted.

Pro-tip: Check out Claude or Gemini. They hallucinate far less on coding tasks. Alternatively, enable internet search on o3 which boosts its ability to reference online documentation and real world usage examples. I get having a bad taste in your mouth but these tools _aren't _ magic and do have something of a steep learning curve in order to get the most out of them. Not dissimilar from vim/emacs (or lots of dev too…

> enable internet search on o3 I didn't know it could even be disabled. It must be enabled by default, right?

You're correct. Tapping the globe icon encourages the model to use it more often.
Post reply on HN