Comments URL: https://news.ycombinator.com/item?id=47071517

Points: 5

# Comments: 0

These kinds of tasks are relatively easy for LLMs, they’re operating in a solved design space and recombining known patterns. It looks impressive to us because writing a compiler from scratch is difficult and time consuming for a human, not because of the problem itself.

That doesn’t mean LLMs aren’t useful, even if progress plateaued tomorrow, they’d still be very valuable tools. But building yet another C compiler or browser isn’t that compelling as a benchmark. The industry keeps making claims about reasoning and general intelligence, but I’d expect to see systems producing genuinely new approaches or clearly better solutions, not just derivations of existing OSS.

Instead of copying a big project, I'd be more impressed if they could innovate in a small one.

It's a perfectly valid benchmark and very telling.

I'm sure this is handy for LLM usage, but this was a problem before those were a thing I'd say.

Youtubers are a couple of magnitudes away from that, AFAIK

But it’s the default call to action for bigger inquiries

It’s a very nice application, easy to configure, very customisable, stable. I’ve only have experienced one bug with app, I opened an issue in GH and Wez had a nightly version with a fix in just a few hours.

If you haven’t check it out yet, I highly recommend it.

Me and my friend started exploiting the bug like crazy, spending hours literally just doing that. The problem was that the process to exploit this bug was very convoluted and a pain to do manually.

I had no experience of programming nor anyone around who knew anything about it, but I felt that there must be a way to do the whole process automatically. Computers are meant to automate stuff, right? Anyway, I started investigating online and found AutoIt, which is a BASIC-like scripting language that allows you to automate GUI stuff.

I then began building a script that did literally what I had to do manually to exploit that bug. And literally means literally. The way I made it work was by scanning colors in coordinates (to check whether a window it's opened, etc), moving the mouse and clicking. The script was full of duplicate code. I didn't knew about loops. All I knew was if conditions and mouse actions. I even remember having to go to a forum and asking for help (in super broken English) because I wanted to keep the program running indefinitely (they told me to use `while true`). But it worked.

Anyway, I spent about two weeks of after school afternoons building this bot. Then, when I had it working, got banned in a matter of days :^)

This might not seems like much, but I remember it fondly as this was my introduction to programming and the reason this became my career. I'm still holding onto the (terrible, terrible) script.

[0]: https://github.com/vercel/satori#html-elements [1]: https://og-playground.vercel.app