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

Building a Personal AI Factory

john-rush.com

Post reply on HN
Print view View original

91–100 of 167 posts

Re: Building a Personal AI Factory

#91

Post by _1tem »

marviel wrote: post #20

Thanks for the writeup! I talked about a similar, but slightly simpler workflow in my post on "Vibe Specs". https://lukebechtel.com/blog/vibe-speccing I use these rules in all my codebases now. They essentially cause the AI to do two things differently: (1) ask me questions first (2) Create a `spec.md` doc, before writing any code. Seems not too dissimilar from yours, but I limit it to a single LLM

Claude Code now handles this natively with “plan mode”. Bit slow and annoying to do it manually with .md files in my opinion.

Re: Building a Personal AI Factory

#92

Post by th0ma5 »

simonw wrote: post #53
low_common wrote: post #46

Earlier quoted context omitted.

That's a pretty trivial example for one of these IDEs to knock out. Assembly is certainly in their training sets, and obviously docker is too. I've watched cursor absolutely run amok when I let it play around in some of my codebase. I'm bullish it'll get there sooner rather than later, but we're not there yet.

I think the hardest problem in computer science right now may be coming up with an LLM demo that doesn't get called "pretty trivial".

Maybe you should try something other than demos? Have you tried creating a reliable system?

Re: Building a Personal AI Factory

#93

Post by mmarian »

And here I am struggling to get Claude to create a nice-looking search bar a la booking.com , with some adjustments for my personal use case; it does ok, but never gets to the end result and once I refreshed my Tailwind knowledge it felt much slower than hand coding. I feel like I'm living in a different world.

Re: Building a Personal AI Factory

#94

Post by stavros »

geekymartian wrote: post #32

ADHD coding, brute forcing product generation until you get it right? Just freaking write the code that you can expand and modify in the future instead of increasing your carbon footprint.

Man, programming has changed forever, and the sooner you realize that, the better for you. Saying "write the code" is like telling people to shoe their own horses instead of dealing with them newfangled cars that can break down.

Re: Building a Personal AI Factory

#95

Post by stavros »

tptacek wrote: post #36
ClawsOnPaws wrote: post #10

Earlier quoted context omitted.

I keep coming to the same conclusion, which basically is: if I had an LLM write it for me, I just don't care about it. There are 2 projects out of the maybe 50 or so that are LLM generated, and even for those two I cared enough to make changes myself without an LLM. The rest just sit there because one day I thought huh wouldn't it be neat if, and then realized actually I cared more about having that thought than havi…

We just had a story last night about a Python cryptography maintainer using Claude to add formally-verified optimizations to LLVM. I think the ship has sailed on skepticism about whether LLMs are going to produce valuable code; you can follow Simon Willison's blog for more examples.

I don't understand people who are sceptical about whether LLMs can give value. We're way past that, now at the stage where we're trying to figure out how to extract the most value out of them, but I guess humans don't like change much.

Re: Building a Personal AI Factory

#96

Post by jkhdigital »

simonw wrote: post #53
low_common wrote: post #46

Earlier quoted context omitted.

That's a pretty trivial example for one of these IDEs to knock out. Assembly is certainly in their training sets, and obviously docker is too. I've watched cursor absolutely run amok when I let it play around in some of my codebase. I'm bullish it'll get there sooner rather than later, but we're not there yet.

I think the hardest problem in computer science right now may be coming up with an LLM demo that doesn't get called "pretty trivial".

No the hardest problem is teaching CS undergrads. I just started this year (no background in academia, just 75% of a PhD and well-rounded life experience) and I’ve basically torn up the entire curriculum they handed to me and started vibe-teaching.

Re: Building a Personal AI Factory

#97

Post by stavros »

petesergeant wrote: post #45
apwell23 wrote: post #37

ppl are getting slowly disillusioned with vibe coding. yes AI assisted workflow might be here to stay but it won't be the magical put programmers out of job thing. And this the best product market fit for LLMs. I imagine it will be even worse in other domains.

> ppl are getting slowly disillusioned with vibe coding. This is the absolute polar opposite from my experience. I'm in a large non-tech community with a coders channel, and every day we get a few more Claude Code converts. I would say that vibe-coding is moving into the main-stream with experienced, professional developers who were deeply skeptical a few months ago. It's no longer fancy auto-complete: I have myself…

I don't doubt that LLMs are extremely useful for making simple things quickly. I haven't been able to get them to write hard code on their own, though. I was trying to make a sound card with a Pi Pico the other day, and had crackling and popping in the audio. I kept telling Opus to fix that, it kept being absolutely convinced it knows what the problem is every time, and went through multiple iterations of being absolutely sure it will solve the problem this time (with every time bringing a different reason for why the pops are there), and spent $35.

In the end, it had written 500 lines, the problem was still there, and the code didn't work any differently. It worries me that I don't know what those 500 lines were for.

In my experience, LLMs are amazing for writing 10-20 lines at a time, while you review and fix any errors. If I let them go to town on my code, I've found that's an expensive way to get broken code.

Re: Building a Personal AI Factory

#98

Post by sussmannbaka »

gerdesj wrote: post #34
simonw wrote: post #28

My hunch is that this article is going to be almost completely impenetrable to people who haven't yet had the "aha" moment with Claude Code. That's the moment when you let "claude --dangerously-skip-permissions" go to work on a difficult problem and watch it crunch away by itself for a couple of minutes running a bewildering array of tools until the problem is fixed. I had it compile, run and debug a Mandelbrot fract…

Crack on - this is YC! Why are you not already a unicorn?

As it turns out, the VC potential of Mandelbrot and HelloWorld.py are quite limited :o)

Re: Building a Personal AI Factory

#99

Post by hamstergene »

mmarian wrote: post #93

And here I am struggling to get Claude to create a nice-looking search bar a la booking.com , with some adjustments for my personal use case; it does ok, but never gets to the end result and once I refreshed my Tailwind knowledge it felt much slower than hand coding. I feel like I'm living in a different world.

I think coding assistants aren't great at UI/UX yet because they can't see, their understanding of left/right/lighter/darker is guessed from textual descriptions that accompanied CSS tutorials but they are never actually imagining the looks of what they are working with. I had Cursor repeatedly fix and mess up a CSS grid, over and over again, until I switched to HTML table so that browser would handle layout. Once switched from visuals ("leftmost") to semantics ("first cell in a row") the agent immediately started getting tasks done right.

I guess keep them on backend/library tasks for now. I am sure the companies are already working on getting a snapshot of a browser page and feeding it back into multimodal model so it can comprehend what "looking" means.

Re: Building a Personal AI Factory

#100

Post by 1dom »

simonw wrote: post #53
low_common wrote: post #46

Earlier quoted context omitted.

That's a pretty trivial example for one of these IDEs to knock out. Assembly is certainly in their training sets, and obviously docker is too. I've watched cursor absolutely run amok when I let it play around in some of my codebase. I'm bullish it'll get there sooner rather than later, but we're not there yet.

I think the hardest problem in computer science right now may be coming up with an LLM demo that doesn't get called "pretty trivial".

I'm very pro LLM and AI. But I completely agree with the comment about how many pieces praising LLMs are doing so with trivial examples. Trivial might not be the right word, but I can't think of a better one that doesn't have a negative connotation, but this shouldn't be negative. Your examples are good and useful, and capture a bunch of tasks a software engineer would do.

I'd say your mandelbrot debug and the LLVM patch are both "trivial" in the same sense: they're discrete, well defined, clear-success-criteria-tasks that could be assigned to any mid/senior software engineer in a relevant domain and they could chip through it in a few weeks.

Don't get me wrong, that's an insane power and capability of LLMs, I agree. But ultimately it's just doing a day job that millions of people can do sleep deprived and hungover.

Non-trivial examples are things that would take a team of different specialist skillsets months to create. One obvious potential reason why there's few non-trivial AI examples is because non-trivial AI examples require non-trivial amount of time to be able to generate and verify.

A non-trivial example isn't an example you can look at the output and say "yup, AI's done well here". It requires someone spends time going into what's been produced, assessing it, essentially redesigning it as a human to figure out all the complexity of a modern non-trivial system to confirm the AI actually did all that stuff correctly.

An in depth audit of a complex software system can take months or even years and is a thorough and tedious task for a human, and the Venn diagrams of humans who are thinking "I want to spend more time doing thorough, tedious code tasks" and "I want to mess around with AI coding" is 2 separate circles.

Post reply on HN