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

I quit. The clankers won

dbushell.com

Post reply on HN
Print view View original

211–220 of 515 posts

Re: I quit. The clankers won

#211

Post by zozbot234 »

coldtea wrote: post #204
zozbot234 wrote: post #189

Earlier quoted context omitted.

> Demonstrably working, as in you can prove the code actually works by then putting it to use. That's not how you prove that code works properly and isn't going to fail due to some obscure or unforessen corner case. You need actual proof that's driven by the code's overall structure. Humans do this at least informally when they code, AI's can't do that with any reliability, especially not for non-trivial projects (fo…

> That's not how you prove that code works properly and isn't going to fail due to some obscure or unforessen corner case. So? We didn't prove human code "isn't going to fail due to some obscure or unforessen corner case" either (aside the tiny niche of formal verification). So from that aspect it's quite similar. > so most coding agents simply work their way iteratively to get their test results to pass. That's not…

> describe the issue exactly in the way a human programmer would

Human programmers don't usually hallucinate things out of thin air, AIs like to do that a whole lot. So no, they aren't working the exact same way.

Re: I quit. The clankers won

#212

Post by oompydoompy74 »

mchaver wrote: post #167
oompydoompy74 wrote: post #71

Good lord I’m going to have to figure out some way to filter Hacker News. I’m so tired of this same sort of article (and the opposite) being posted every day. AI isn’t going away. AI is better than you think it is. AI is probably also worse than you think it is. The world has nuance, so can we please all chill?

These conversations can add to the nuance. Anyway, you can just vibe code the filter you want and be done with it.

I disagree that this adds anything new to the conversation, but fair enough.

Re: I quit. The clankers won

#213

Post by anticorporate »

pfisherman wrote: post #131
Waterluvian wrote: post #44

Improving developer skills is not valuable to your company. They don't tell a customer how many person-hours of engineering talent improvement their contract is responsible for. They just want a solved problem. Some companies comprehend how short-sighted this is and invest in professional development in one way or another. They want better engineers so that their operations run better. It's an investment and arguably…

This is going to catch some heat, but what if the most important professional “developer skill” to learn or improve is how to effectively use coding agents? I saw something similar in ML when neural nets came around. The whole “stack moar layerz” thing is a meme, but it was a real sentiment about newer entrants into the field not learning anything about ML theory or best practices. As it turns out, neural nets “won”…

I'd offer an edit that the most important skill may be knowing when the agent is wrong.

There's so much hand wringing about people not understanding how LLMs work and not nearly enough hand wringing about people not understanding how computer systems work.

Re: I quit. The clankers won

#214

Post by bragr »

theshrike79 wrote: post #104
tuvang wrote: post #97

Earlier quoted context omitted.

There is a deadly game of chicken going on. Junior recruiting already stopped for the most part. Only way this doesn’t end in a catastrophe is if AI becomes genuinely as good as the most skilled developers before we run out of them. Which I doubt very much but don’t find completely impossible.

And the irony is that AI usage should make onboarding juniors easier. Before it was "hey $senior_programmer where's the $thing defined in this project?", which either required a dedicated person onboarding or someone's flow was interrupted - an expected cost of bringing up juniors. Now a properly configured AI Agent can answer that question in 60 seconds, unblocking the Junior to work on something. And no, it doesn't…

The problem is that juniors given access to AI don't seem to learn as much. AI just gives them fish over and over instead of learning how to fish.

Re: I quit. The clankers won

#215

Post by beej71 »

staminade wrote: post #95

Anti-AI articles like this seem to be the new "Doing my part to resist big tech: Why I'm switching back from Chrome to Firefox" genre that popped up on HN for a decade or so. If it makes you feel better, great, but don't kid yourself that your actions will make any difference whatsoever to the overall trajectory of AI adoption in IT or society.

I love it if it would affect the trajectory, but I don't think it will. I do think it will affect my trajectory, though.

Re: I quit. The clankers won

#216

Post by agentultra »

nicksergeant wrote: post #72
titzer wrote: post #59

Earlier quoted context omitted.

The irony is that the vast deskilling that's happening because of this means that most "software engineers" will become incapable of understanding, let alone fixing or even building new versions of the systems that they are utterly dependent on. There should be thousands or tens of thousands people worldwide that can build the operating systems, virtual machines, libraries, containers, and applications that AI is bui…

I feel I've upskilled in so many directions (not just "ability to prompt LLMs") since going all in on LLM coding. So many tools, techniques, systems, and new areas of research I'd never have had the time to fully learn in the past. I have a hard time believing any tenured developer is not actually learning things when using LLMs to build. They make interesting choices that are repeatable (new CLIs I didn't even know…

> Anyone not learning things via LLM coding right now either doesn't care at all about the underlying code/systems

How many bytes is a pointer in C? How many bytes is a shared pointer in C++? What does sysctl do? What about fsync?

What is a mutex lock? How is it different from a spin lock?

You want to find the n nearest points to a given point on a 2-D Cartesian plane. Could you write the code to solve that on your own?

Can you answer any of these questions without searching for the answer?

I don't use LLMs and I learn things fine. Always have. For several decades. I care deeply about the underlying code and systems. It annoys me when people say they do and they cannot even understand how the computer works. I'm fine with people having domain-specific knowledge of programming: maybe you've only been interested in web development and scripting DOM elements. But don't pretend that your expertise in that area means you understand how to write an operating system.

Or worse: that it prevents you from learning how to write an operating system.

You can do that without an LLM. There's no royal road. You have to understand the theory, read the books, read the code, write the code, make mistakes, fix mistakes, read papers, talk to other people with more experience than you... and just write code. And rewrite it. And do it all again.

I find the opposite is true: those who use LLM coding exclusively never enjoyed programming to begin with, only learned as much as they needed to, and want the end results.

Re: I quit. The clankers won

#217

Post by kstenerud »

simoncion wrote: post #181
Supermancho wrote: post #108

Earlier quoted context omitted.

> That is basically your argument. Like AI is a copyright theft machine, with companies owning the entire stack and being able to take away at will, and comitting crimes like decompiling source code instead of clean room is not a selling point either... Stop trying to make this into some abstract argument. It's not an argument anymore. It's already happened. How one might choose to characterize the reality, is irrele…

> It's already happened. Agreed. > Stop trying to make this into some abstract argument. As you mentioned, it's not an abstract argument. It's statements of fact. > A vast (and growing) amount of source code is more open... No, not at all. 1) If you honestly believe that major tech companies will permit both copyright- and license-washing of their most important proprietary code simply because someone ran it through…

This is looking at the current situation through the old lens.

That ship has sailed. The revolution is happening. We live in a new reality now, one where we're still trying to figure out what rules should even be.

And there will be winners and losers, and copyright and patent law will be modified in an attempt to tame the chaos, with mixed results because of all of the powerful players on both ends.

You can live on the front of it for high risk/reward, or at the back for safety. But either way, you're going to exist in this new reality and you need to decide your risk appetite.

Re: I quit. The clankers won

#219

Post by rekrsiv »

pfisherman wrote: post #131
Waterluvian wrote: post #44

Improving developer skills is not valuable to your company. They don't tell a customer how many person-hours of engineering talent improvement their contract is responsible for. They just want a solved problem. Some companies comprehend how short-sighted this is and invest in professional development in one way or another. They want better engineers so that their operations run better. It's an investment and arguably…

This is going to catch some heat, but what if the most important professional “developer skill” to learn or improve is how to effectively use coding agents? I saw something similar in ML when neural nets came around. The whole “stack moar layerz” thing is a meme, but it was a real sentiment about newer entrants into the field not learning anything about ML theory or best practices. As it turns out, neural nets “won”…

The endgame in programming is reducing complexity before the codebase becomes impossible to reason about. This is not a solved problem, and most codebases the LLMs were trained on are either just before that phase transition or well past it.

Complexity is not just a matter of reducing the complexity of the code, it's also a matter of reducing the complexity of the problem. A programmer can do the former alone with the code, but the latter can only be done during a frank discussion with stakeholders.

A vibe coder using an LLM to generate complexity will not be able to tell which complexity to get rid of, and we don't have enough training data of well-curated complexity for LLMs to figure it out yet.

Re: I quit. The clankers won

#220

Post by niek_pas »

malwrar wrote: post #47

I do find it hard to tolerate the feeling of being watched online. The second-most trending dataset on huggingface right now is a snapshot of HN updating at a 5 minute interval. It makes me not want to really comment at all, just like how I don’t really publish any software I write anymore. Turns out it sucks to produce original works when you know that, whereas previously a few people at best might see your work, no…

This is really interesting to me, because it never occurred to me to feel this way. Why would I care whether my comments are ending up in some dataset somewhere that's being used to train some model? My comments are boring and mostly uninformed. Have at it.

I'm curious: would you say the feeling of being watched online is making you afraid of some repercussion, or is it something else?

Post reply on HN