Live data from Hacker News

Is anybody else bored of talking about AI?

blog.jakesaunders.dev

531–540 of 570 posts

Re: Is anybody else bored of talking about AI?

#531
post #459
post #406

Earlier quoted context omitted.

The productivity comes from not having the startup costs. You don’t need to research the best way to do X, just verify that X works via tests and documentation. I find it still takes T hours to actually implement X with an agent, but if I didn’t know how to do X it eliminates that startup cost which might make it take 3T hours instead. The only downside is not learning about method Y or Z that work differently than X…

> You don’t need to research the best way to do X, just verify that X works via tests and documentation. "Just verify" is glossing over a lot of difficult work, though. It doesn't just involve checking whether the program compiles and does what you wanted—that's the easy part. You should also verify that the program is secure, robust, reasonably performant, efficient, etc. Even if you think about these things, and as…

Exactly. "Tests pass" and "code is secure" are just different things. AI code makes that gap worse.

I run static analysis on mixed human/AI codebases. The AI parts pass tests fine but they'll have stuff any SAST tool flags on first run — hardcoded creds, wildcard CORS, string-built SQL. Works in a demo, turns into a CVE in prod.

And nobody's review capacity scaled with generation speed. Most teams don't even have semgrep in CI. So you get unreviewed code just sitting in production.

The "10x" is real if you count lines shipped. Nobody counts the fix cost downstream though.

Re: Is anybody else bored of talking about AI?

#532

Earlier quoted context omitted.

Fuck, another C developer who will just let memory leaks grow uncontrollably is born.

You might wanna learn how to set up constraints for your agents. Every memory map is accounted for with docs pointing to the exact structs, allocations, and usage patterns. It's stuff I was already doing. Now I can do it in a fraction of the time.

So you have absolutely NO doubt

Re: Is anybody else bored of talking about AI?

#534

Earlier quoted context omitted.

Exactly. People love to compare LLMs to power tools for carpenters and smiths. But if my miter saw had a 20% chance to produce cuts at a 45 degree angle when I have it set for 90, I would throw it out so fast I would leave Looney Tunes style tracks. A tool which only sometimes does its job is worse than no tool at all.

To be rather pedantic your miter saw probably doesn't cut exactly 90 degrees. Especially if you reset it. LLMS are low accuracy for sure but so are humans. I am not saying AI is going to replace us all in a whole entirety my broader point is that these tools will be another tool that changes the market share of jobs.

The low accuracy of human guesstimation is why we use deterministic tools, not so-called tools that imitate our ability (or worse).

Tools are not replacements for people! Tools are enhancements.

AI is an attempt to replace people with something unhuman.

Re: Is anybody else bored of talking about AI?

#535
post #347

Earlier quoted context omitted.

This is really not true. There are stories of people who had no background in software engineering who now write entire applications using AI. And I have personally seen this happen.

Before AI, there were also stories of people who had no background in software engineering who wrote entire applications using their fingers. This was called "learning to be a software engineer". I don't mean to snipe at AI, because it really does seem to have set more people on the path of learning, but I was writing VB5 apps when I was 14 by copying poorly understood bits and pieces from books. Now people are doing…

Learning to code and write an application was very hard for most people, because of time and other friction. I know of non-coders who now have applications running for various things they find useful.

You might not consider this productive, but they do, so what you think literally doesn't matter to them.

Re: Is anybody else bored of talking about AI?

#536
post #22

It's a black box that thinks for me, sometimes it's good, sometimes it's bad, sometimes it times out. I am extremely skeptical of AI products anyone builds. It's just using one black box to build scaffolding around another black box and then typically want to charge money for it. I don't see any value there.

A horse without gear is a wild animal. Slap on a saddle, some reigns, and training and it’s suddenly a transport vehicle.

AI products can and do help make the raw models applicable to targeted domains. Think of them as a black box sure, but that doesn’t mean they dont add value.

Re: Is anybody else bored of talking about AI?

#537

Earlier quoted context omitted.

You might wanna learn how to set up constraints for your agents. Every memory map is accounted for with docs pointing to the exact structs, allocations, and usage patterns. It's stuff I was already doing. Now I can do it in a fraction of the time.

So you have absolutely NO doubt

Doubts? I'm full of them. I'm building a new task manager for myself that has things I've always wanted. It's basically a profiler so it must be performant.

https://github.com/hparadiz/evemon

The core if you disable all the plugins is currently topped out at 73.8 MB after several days of running it. I've given it several audits with the AI agents using actual memory maps and doing the math on each variable.

I haven't had time to do Milkdrop yet but it's on my todo. The issue isn't doing the work. The issue is not having enough credits in my accounts to throw some compute at it. I'll get to it eventually. But it's actually way easier now to try new ways of packing the data into binary and profiling it for issues.

The issues I've had are edge cases like a 6 hour youtube stream. At one point the BPM detector was buffering the entire track in the pipewire sink. It took one throwaway prompt to the AI to solve that one.

Re: Is anybody else bored of talking about AI?

#538

Earlier quoted context omitted.

> So some skills fade and others rise. And also, software has moved in predictable cycles for many decades at this point. We are still a very young field but we do have some history at this point. And there'll be a split too... like there's a giant divide between those mechanics who used to work on carburetors and the new gen with microcontrollers, injection systems, etc. People who think cars are 'too complicated' a…

And to take the analogy even further, I'm sure there will be a subset of people who develop really strong opinions about a particular toolchain or workflow. Like how we have people who specialize in 70s diesel trucks or 90-00s JDM sports cars, there'll likely be programmers who are SMEs at updating COBAL to Rust using Claude.

yup. I do early bosch EFI and K-Jet... thats what I know!

Re: Is anybody else bored of talking about AI?

#539

I’m sad that it’s crowded out all the interesting stuff I used to love learning about on HN.

HN needs an "AI" tab, and the ability to flag things as AI

Some of it very interesting, but maybe it shouldn't be on the home page unless it's a certain critical mass (similar to show HN)?

Re: Is anybody else bored of talking about AI?

#540
post #219

Earlier quoted context omitted.

Well, on the other hand, software isn’t all about checking emails. I know someone who worked for a nonprofit that made pregnancy health software that worked over text messaging. Its clients were women in Africa who didn’t have much, but they had a cell phone, so they could get reminders, track vitals, and so forth. They had to find enough funding to pay several software engineers to build and maintain that system. If…

So, here is a case where AI was a technical fix to a social problem: that as a global society, funding is skewed away from things which benefit people. Google can find the budget for as many ad-touchers as it wants, but nonprofits have to scrounge and make do by paying for "metered intelligence" from a megacorp. So in isolation, I think it's great that they managed to achieve this. But I mourn that the only way they…

To be clear, this work was done years pre-AI. I was just giving an example of how lowering software development costs could be a public good.
Post reply on HN