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

I don't care how well your "AI" works

fokus.cool

Post reply on HN
Print view View original

481–490 of 808 posts

Re: I don't care how well your "AI" works

#481

Post by idiotsecant »

robot-wrangler wrote: post #361
mattgreenrocks wrote: post #311

Earlier quoted context omitted.

> This is the culture that replaced hacker culture. Somewhere along the lines of "everybody can code," we threw out the values and aesthetics that attracted people in the first place. What began as a rejection of externally imposed values devolved into a mouthpiece of the current powers and principalities. This is evidenced by the new set of hacker values being almost purely performative when compared against the old…

I realized recently that if you want to talk about interesting topics with smart people, if you expect things like critical thinking and nuanced discussion, you're currently much better off talking literature or philosophy than anything related to tech. I mean, everyone knows that discussing politics/economics is rather hopelessly polarized, everyone has their grievances or their superstitions or injuries that they c…

This is classically framed as philosophy vs sophistry. The truth is that both are necessary, but only one makes money. When your entire culture assigns value with money it's obvious which way the scales will tip.

Re: I don't care how well your "AI" works

#482

Post by Glemkloksdjf »

HWR_14 wrote: post #465
JustExAWS wrote: post #430

Earlier quoted context omitted.

I’ve been a “software engineer” or closely adjacent for 30 years. During that time, I’ve worked for small and medium “lifestyle companies”, startups, boring Big Enterprise, $BigTech and over the past 5 years (including my time at $BigTech) worked as a customer facing cloud consultant where I’ve seen every type of organization imaginable and how they work. No one ever gave a rip about “craftsmanship”. They hire you fo…

> If an LLM helps me exchange my labor for money more efficiently Except that's unproven. It might make you more productive, but whether you get any of that new value is untested.

Right now its just a tool you can use or not and if you are smart enough, you figure out very quickly when to use a tool for efficency and when not.

I do not vibe code my core architecture because i control it and know it very well. I vibe code some webui i don't care about or a hobby idea in 1-4h on a weekend because otherwise it would take me 2 full weekends.

I fix emails, i get feedback etc.

When I do experiemnts with vibe coding, i'm very aware what i'm doing.

Nonetheless, its 2025. Alone 2026 we will add so much more compute and the progress we see is just crazy fast. In a few month there will be the next version of claude, gpt, gemini and co.

And this progress will not stop tomorrow. We don't know yet how fast it will progress and when it will be suddenly a lot better then we are.

Additionally you do need to learn how to use these tools. I learned through vibe coding that i have to specify specific things i just assume the smart LLM will do right without me telling for example.

Now i'm thinking about doing an experiemnt were i record everything about a small project i want to do, to then subscribe it into text and then feeding it into an llm to strucuture it and then build me that thing. I could walk around outside with a headset to do so and it would be a fun experiemnt how it would feel like.

I can imagine myself having some non intrusive AR Google and the ai sometimes shows me results and i basically just give feedback .

Re: I don't care how well your "AI" works

#483

Post by pixl97 »

idiotsecant wrote: post #476
everdrive wrote: post #442

Earlier quoted context omitted.

>Somewhere along the lines of "everybody can code," we threw out the values and aesthetics that attracted people in the first place. At some point people started universally accepting the idea that any sort of gatekeeping was a bad thing.I think by now people are starting to realize that this was a flawed idea. (at best, gatekeeping is not a pure negative; it's situational) But, despite coming to realize this I think…

Your point is that the hacker ethos involved ... Fewer people being excited about programming? I don't think we experienced this on the same planet. Web 1.0 was full of weirdos doing cool weird stuff for the pure joy of discovery. That's the ethos we need back, and it's not incompatible with AI. The wrong turn we took was letting business overtake joy. That's a decision we can undo today by opting out of that whole e…

> That's a decision we can undo today by opting out of that whole ecosystem.

Ah yes, we'll also skip out on eating too.

Re: I don't care how well your "AI" works

#484

Post by CamperBob2 »

mattgreenrocks wrote: post #311
easterncalculus wrote: post #244

> I find it particularly disillusioning to realize how deep the LLM brainworm is able to eat itself even into progressive hacker circles. That's the thing, hacker circles didn't always have this 'progressive' luddite mentality. This is the culture that replaced hacker culture. I don't like AI, generally. I am skeptical of corporate influence, I doubt AI 2027 and so-called 'AGI'. I'm certain we'll be "five years away"…

> This is the culture that replaced hacker culture. Somewhere along the lines of "everybody can code," we threw out the values and aesthetics that attracted people in the first place. What began as a rejection of externally imposed values devolved into a mouthpiece of the current powers and principalities. This is evidenced by the new set of hacker values being almost purely performative when compared against the old…

We need to talk seriously and plainly about the spiritual and existential damage done by LLMs.

I'm tempted to say "You're not helping," as my eyes roll back in their sockets far enough to hurt. But I can also understand how threatening LLMs must appear to programmers, writers, and artists who aren't very good at their jobs.

What I don't get is why I should care.

Re: I don't care how well your "AI" works

#485

Post by thunky »

vladimirralev wrote: post #189
spicyusername wrote: post #115

Earlier quoted context omitted.

> assign work to an LLM This is just not happening anywhere around me. I don't know why it keeps getting repeated in every one of these discussions. Every software engineer I know is using LLM tools, but every team around me is still hiring new developers. Zero firing is happening in any circle near me due to LLMs. LLMs can not do unsupervised work, period. They do not replace developers. They replace Stack Overflow…

Today's high-end LLMs can do a lot of unsupervised work. Debug iterations are at least junior level. Audio and visual output verification is still very week (i.e. to verify web page layout and component reactivity). Once the visual model is good enough to look at the screen pixels and understand, it will instantly replace junior devs. Currently if you have only text output all new LLMs can iterate flawlessly and solv…

> Once the visual model is good enough to look at the screen pixels and understand, it will instantly replace junior devs

Curious if you gave Antigravity a try yet? It auto-launches a browser and you can watch it move the mouse and click around. It's able to review what it sees and iterate or report success according to your specs. It takes screen recordings and saves them as an artifact for you to verify.

I only tried some simple things with it so far but it worked well.

Re: I don't care how well your "AI" works

#486

Post by d_burfoot »

> AI systems exist to reinforce and strengthen existing structures of power and violence. They are the wet dream of capitalists and fascists.

Persuasion tip: if you write comments like this, you are going to immediately alienate a large portion of your audience who might otherwise agree with you.

Re: I don't care how well your "AI" works

#487

Post by JustExAWS »

HWR_14 wrote: post #465
JustExAWS wrote: post #430

Earlier quoted context omitted.

I’ve been a “software engineer” or closely adjacent for 30 years. During that time, I’ve worked for small and medium “lifestyle companies”, startups, boring Big Enterprise, $BigTech and over the past 5 years (including my time at $BigTech) worked as a customer facing cloud consultant where I’ve seen every type of organization imaginable and how they work. No one ever gave a rip about “craftsmanship”. They hire you fo…

> If an LLM helps me exchange my labor for money more efficiently Except that's unproven. It might make you more productive, but whether you get any of that new value is untested.

Well I have personally tested it on the green field projects I mostly work on and it does the grunt work of IAC (Terraform) and even did a decently complicated API with some detailed instructions like I would give another developer.

I’ve done literally dozens of short term quick turn around POCs from doing the full stack from an empty AWS account to “DevOps” to the software development -> training customers how to fish and showing them the concepts -> move on to next projects between working at AWS ProServe and now a third party consulting company. I’m familiar with the level of effort for these types of projects. I know how many fewer man hours it takes me now.

I have avoided front end work for well over a decade. I had to modify the front end part of the project we released to the customer that another developer did to remove all of the company specific stuff to make it generic so I could put it in our internal repo. I didn’t touch one line of front end code to make the decently extensive modifications, honestly I didn’t even look at the front end changes. I just made sure it worked as expected.

Re: I don't care how well your "AI" works

#488

Post by danillonunes »

eisbaw wrote: post #122

Programming and CS is the art of solving problems - hopefully problems that matter. AI lets you do that faster. AI may suggest a dumb way, so you have to think, and tell it what to do. My rate of thinking is faster than typing, so the bottleneck has switched from typing to thinking! Don't let AI think for you. Do actual intensional arch design. Programmers that don't know CS who only care about hammering the keyboard…

> My rate of thinking is faster than typing, so the bottleneck has switched from typing to thinking!

Unless you're neuralinking to AI, you're still typing.

What changed is what you type. You type less words to solve your problem. The machine does the conversion from less words to more words. At the expense of less precision: the machine can do the conversion to the incorrect sequence of more words.

Re: I don't care how well your "AI" works

#489

Post by fragmede »

terminalbraid wrote: post #444
MrDarcy wrote: post #420

Earlier quoted context omitted.

I’m a hacker and I’d show respect to that latter person if they did the subcontracting and reviewed their craft well.

But you wouldn't call them a craftsperson because they didn't do any craft other than "be a manager". Reviewing work is not on the same plane as actually creating something.

Does Steve Jobs deserve any respect for building the iPhone then? What is this "actually creating"? I'm sure he wasn't the one to do any of the "actually creating" and yet, there's no doubt in my mind that he deserves credit for the iPhone existing and changing the world.

Re: I don't care how well your "AI" works

#490

Post by Wowfunhappy »

thewebguyd wrote: post #462
easterncalculus wrote: post #244

> I find it particularly disillusioning to realize how deep the LLM brainworm is able to eat itself even into progressive hacker circles. That's the thing, hacker circles didn't always have this 'progressive' luddite mentality. This is the culture that replaced hacker culture. I don't like AI, generally. I am skeptical of corporate influence, I doubt AI 2027 and so-called 'AGI'. I'm certain we'll be "five years away"…

> All that said, the actual workday is absolutely filled with busy work that no one really wants to do, and the refusal of a loud minority to engage with that fact is what's leading to this. The attitude and push back from this loud minority has always been weird to me. Ever since I got my hands on my first computer as a kid, I've been outsourcing parts of my brain to computing so that I can focus on more interesting…

> Ever since I got my hands on my first computer as a kid, I've been outsourcing parts of my brain to computing so that I can focus on more interesting things. I no longer have to remember phone numbers, I no longer have to carry a paper notepad, my bookshelf full of reference books that constantly needed to be refreshed became a Google search away instead. Intellisense/code completion meant I didn't have to waste time memorizing every specific syntax and keyword. Hell, IDEs have been generating code for a long time. I was using Visual Studio to automatically generate model classes from my database schema for as long as I can remember, and even generating CRUD pages.

I absolutely agree with you, but I do think there's a difference in kind between a deterministic automation you can learn to use and get better at, and a semi-random coding agent.

The thing I'm really struggling with is that unlike e.g. code completion, there doesn't seem to be a clear class of tasks that LLMs are good at vs bad at. So until the LLMs can do everything, how do I keep myself in the loop enough that I'll have the requisite knowledge to step in when the LLM fails?

You mention how technology means we no longer have to remember phone numbers. But what if all digital contact lists had a very low chance of randomly deleting individual contacts over time? Do you keep memorizing phone numbers? I'm not sure!

Post reply on HN