[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: 1763963706245605.jpg (47 KB, 600x600)
47 KB JPG
My fellow programmers.
In my estimation, we can compete with LLMs on quality for the foreseeable future.
However, nobody cares about quality. We must also compete with LLMs on speed.
How can we do that? What technology could help us catch up?
>>
>>108967063
>However, nobody cares about quality.
Don't code slop and they'll care
>>
this snailcat cope thread is so bad it's quite probably bait

have your bump, your (You)... well meme'd, my fren.
>>
>>108967085
The difference in quality is not large enough to solely drive usage decisions at this point.
It would be, if human-written software could match LLM-written slopware on features.
But I am starting to doubt that this is possible in the time we have, with the technology in common use today.
>>
>>108967063
Coding is Snailcat and trans activity
>>
>>108967063
We''ll never compete against LLMs on speed alone, quality is what makes the difference.
Of course, there are timesinks which could be avoided. Use libraries (where appropriate) instead of reinventing the wheel. Don't use memelangs like C or Haskell. No useless features.
>>
>>108967063
meth
>>
I just finished my internship and the company literally forced me to use Claude. I wish I would have gone into networking instead of programming lol. I'm looking for a job in tech support or something programming is dead.
>>
>>108967291
What isn't a memelang?
>>
>>108967307
I am really not sure why all corpos are using claude. Sure it's good but I havent heard a single thing about any company trying openweight models. does everyone really just have so much money to burn
>>
>>108967348
Java, Javascript, C++, C#, Python, Lua
Those are just ones I've used but there's obviously more.
Memelangs: Clojure, C, Haskell, Zig, Visual Basic (rip)
>>
>>108967373
Forcing your reports to use anything other than the trendiest LLM would open you up to attacks from other managers (or at the highest level, other CEOs).
>>
>>108967378
That's funny, Zig programmers strike me as unusually productive compared to anyone using any of those languages. They might have a fighting chance. I doubt anyone writing Java will.
>>
>>108967085
This. "Ship lots of shitty features fast and dirty" devs are going to get BTFO by current AI. MIT-level Software engineers aren't going to get replaced until AI can actually think.
>>
>>108967063
>mwf I must slop shit to work at the vibed codebase with 600k loc / month
>no documentation, no onboarding
>just ask Claudie, bro
>Claudie is shit
>Don't give a fuck now, 0 domain knowledge, just tell ai to do task, track x4 time and push
>If something breaks, I will not know where or give a shit really
Oh, they'll catch to the "quality actually good" this year, anon, you can be sure of it. Speedy slopped shit is unmaintainable.
>>
>>108967063
speed
>>
>>108969141
I’m doing my part as well, just abusing the shit out of Claude and not even looking at the output. I don’t even think my peers reviewing my PRs are even looking at them anymore
>>
AI is just a bunch of other peoples code from github from 2015. I think if you build up a back log of good templates, libraries, and code completion snippets while doing your job I think you could hypothetically compete on speed.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.