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

Andrej Karpathy: Software in the era of AI [video]

youtube.com

Post reply on HN
Print view View original

361–370 of 827 posts

Re: Andrej Karpathy: Software in the era of AI [video]

#361

Post by tinyhouse »

After Cursor is sold for $3B, they should transfer Karpathy 20%. (it also went viral before thanks to him tweeting about it)

Great talk like always. I actually disagree on a few things with him. When he said "why would you go to ChatGPT and copy / paste, it makes much more sense to use a GUI that is integrated to your code such as Cursor".

Cursor and the like take a lot of the control from the user. If you optimize for speed then use Cursor. But if you optimize for balance of speed, control, and correctness, then using Cursor might not be the best solution, esp if you're not an expert of how to use it.

It seems that Karpathy is mainly writing small apps these days, he's not working on large production systems where you cannot vibe code your way through (not yet at least)

Re: Andrej Karpathy: Software in the era of AI [video]

#362

Post by dapperdrake »

eitally wrote: post #347

It's going to be very interesting to see how things evolve in enterprise IT, especially but not exclusively in regulated industries. As more SaaS services are at least partly vibe coded, how are CIOs going to understand and mitigate risk? As more internal developers are using LLM-powered coding interfaces and become less clear on exactly how their resulting code works, how will that codebase be maintained and increme…

Just like when all regulated industries started only using decision trees and ordinary least-squares regression instead of any other models.

Re: Andrej Karpathy: Software in the era of AI [video]

#363

Post by bryanh »

greybox wrote: post #353

He's talking about "LLM Utility companies going down and the world becoming dumber" as a sign of humanity's progress. This if anything should be a huge red flag

Replace with "Water Utility going down and the world becoming less sanitary", etc. Still a red flag?

Re: Andrej Karpathy: Software in the era of AI [video]

#364

Post by flumpcakes »

singularity2001 wrote: post #88
flumpcakes wrote: post #85

Earlier quoted context omitted.

I'm not sure this makes sense as a question. Registers are 'controlled' by running code for a given state. An AI can write code that changes registers, as all code does in operation. An AI can't directly 'control registers' in any other way, just as you or I can't.

what he means is why are the tokens not directly machine code tokens

What is meant by a 'machine code token'? Ultimately a processor needs assembly code as input to do anything. Registers are set by assembly. Data is read by assembly. Hardware is managed through assembly (for example by setting bits in memory). Either I have a complete misunderstanding on what this thread is talking about, or others are commenting with some fundamental assumptions that aren't correct.

Re: Andrej Karpathy: Software in the era of AI [video]

#365

Post by poorcedural »

blobbers wrote: post #114

Software 3.0 is the code generated by the machine, not the prompts that generated it. The prompts don't even yield the same output; there is randomness. The new software world is the massive amount of code that will be burped out by these agents, and it should quickly dwarf the human output.

It is not the code, which just like prompts is a written language. Software 3.0 will be branches of behaviors, by the software and by the users all documented in a feedback loop. The best behaviors will be merged by users and the best will become the new HEAD. Underneath it all will be machine code for the hardware, but it will be the results that dictate progress.

Re: Andrej Karpathy: Software in the era of AI [video]

#366

Post by xnx »

beacon294 wrote: post #315

What is this "clerk" library he used at this timestamp to tell him what to do? https://youtu.be/LCEmiRjPEtQ?si=XaC-oOMUxXp0DRU0&t=1991 Gemini found it via screenshot or context: https://clerk.com/ This is what he used for login on MenuGen: https://karpathy.bearblog.dev/vibe-coding-menugen/

That blog post is a great illustration that most of the complexity/difficulty of a web app is in the hosting and not in the useful code.

Re: Andrej Karpathy: Software in the era of AI [video]

#367

Post by greybox »

bryanh wrote: post #363
greybox wrote: post #353

He's talking about "LLM Utility companies going down and the world becoming dumber" as a sign of humanity's progress. This if anything should be a huge red flag

Replace with "Water Utility going down and the world becoming less sanitary", etc. Still a red flag?

You're making leap of logic.

Before water sanitization technology we had no way of sanitizing water on a large scale.

Before LLMs, we could still write software. Arguably we were collectively better at it.

Re: Andrej Karpathy: Software in the era of AI [video]

#368

Post by layer8 »

bonoboTP wrote: post #265
latexr wrote: post #254

Earlier quoted context omitted.

> Or generating 4 versions of PR for the same task so that the human could just pick the best one. That sounds awful. A truly terrible and demotivating way to work and produce anything of real quality. Why are we doing this to ourselves and embracing it? A few years ago, it would have been seen as a joke to say “the future of software development will be to have a million monkey interns banging on one million keyboar…

If it's monkeylike quality and you need a million tries, it's shit. It you need four tries and one of those is top-tier professional programmer quality, then it's good.

The problem is, for any change, you have to understand the existing code base to assess the quality of the change in the four tries. This means, you aren’t relieved from being familiar with the code and reviewing everything. For many developers this review-only work style isn’t an exciting prospect.

And it will remain that way until you can delegate development tasks to AI with a 99+% success rate so that you don’t have to review their output and understand the code base anymore. At which point developers will become truly obsolete.

Re: Andrej Karpathy: Software in the era of AI [video]

#369

Post by r2b2 »

eitally wrote: post #347

It's going to be very interesting to see how things evolve in enterprise IT, especially but not exclusively in regulated industries. As more SaaS services are at least partly vibe coded, how are CIOs going to understand and mitigate risk? As more internal developers are using LLM-powered coding interfaces and become less clear on exactly how their resulting code works, how will that codebase be maintained and increme…

I've found that as LLMs improve, some of their bugs become increasingly slippery - I think of it as the uncanny valley of code.

Put another way, when I cause bugs, they are often glaring (more typos, fewer logic mistakes). Plus, as the author it's often straightforward to debug since you already have a deep sense for how the code works - you lived through it.

So far, using LLMs has downgraded my productivity. The bugs LLMs introduce are often subtle logical errors, yet "working" code. These errors are especially hard to debug when you didn't write the code yourself — now you have to learn the code as if you wrote it anyway.

I also find it more stressful deploying LLM code. I know in my bones how carefully I write code, due to a decade of roughly "one non critical bug per 10k lines" that keeps me asleep at night. The quality of LLM code can be quite chaotic.

That said, I'm not holding my breath. I expect this to all flip someday, with an LLM becoming a better and more stable coder than I am, so I guess I will keep working with them to make sure I'm proficient when that day comes.

Re: Andrej Karpathy: Software in the era of AI [video]

#370

Post by mirrorlake »

hombre_fatal wrote: post #248
mentalgear wrote: post #157

Meanwhile, I asked this morning Claude 4 to write a simple EXIF normalizer. After two rounds of prompting it to double-check its code, I still had to point out that it makes no sense to load the entire image for re-orientating if the EXIF orientation is fine in the first place. Vibe vs reality, and anyone actually working in the space daily can attest how brittle these systems are. Maybe this changes in SWE with more…

On the other hand, posts like this are like watching someone writing ask jeeves search queries into google 20 years ago and then gesturing how google sucks while everyone else in the room has figured out how to be productive with it and cringes at his "boomer" queries. If you're still struggling to make LLMs useful for you by now, you should probably ask someone. Don't let other noobs on HN +1'ing you hold you back.

Perhaps consider making some tutorials, then, and share your wealth of knowledge rather than calling people stupid.
Post reply on HN