Vibe coding complex changes in Rust
Channel:Jon Gjengset · Length:2:59:07 · Published:2025-08-01 · Views:56125
Summary
Jon Gjengset spends about three hours livecoding complex changes in a Rust codebase with agentic AI tools, testing his hypothesis that AI is weak at non-pattern work. He documents where Claude Code helps and where it struggles.
Why HN recommends it
Pointed to as a non-promotional, realistic long session of an experienced engineer working with AI on non-trivial code.
I found this one and sat down and watched the whole three hours. — HN, 2026
C, C++, Rust, Go & systems languages · Software practice, open source & careers · LLMs & generative AI
Subtitles (excerpt)
Hi everyone, welcome back to another Rust stream. This time we're going to tackle something new. We're still going to be doing Rust, but we're going to try to do vibe coding. Now the caveat to this is I have never done vibe coding before. I've like watched some other people do it. I have installed ClaudeCode and I've set it up with a particular repository. And part of this is because I'm sort of a Luddite when it comes to like a lot of the the current ai let's call it hype whether it's deserved or not um in in particular i have this uh deep-seated fear is the wrong word but perception that these machine learning models don't really understand what they're doing that they are something that they're really good at they are really good at pattern recognition pattern replication merging patterns and being able to reproduce things that seem like they coherently make sense, as in they look reasonable, without necessarily understanding the constructs that happen behind the scenes. What that means is, I think, if I were to try to guess the kinds of things that it's good at, it would be things like writing a WordPress template or writing yet another command line application or talking to the AWS SDK
On Hacker News
Ironically, when machine learning is getting “deeper & deeper”, human learning is getting more and more impatient and shallow. I have been searching “Vibe coding” videos on YouTube that are not promoting something. And I found this one and sat down and watched the whole three hours. It does take a lot of real effort. [VIDEO] — ontouchstart, 2026-01
Jon Gjengset (jonhoo) Who is famously fastidious did a stream on live coding where he did something similar in terms of control. Worth of a watch if that is a style you want to explore. [VIDEO] I don't have the energy to do that for most things I am writing these days which are small PoC where the vibe is fine. I suspect as you do more, you will create dev guides and testing guides that can encap… — sitkack, 2025-08
This video [0] is relevant, though it actually supports your point - it shows Claude Code struggling with non-trivial tasks and needing significant hand-holding. I suspect videos meeting your criteria are rare because most AI coding demos either cherry-pick simple problems or skip the messy reality of maintaining real codebases. [0] [VIDEO] — coffeeri, 2025-09
take a look at the replies to [LINK] eg [LINK] and [VIDEO] PS, what was your last setup? performance depends a lot on the domain/model/workflow[1]; some combinations don't work at all [1] by workflow I mean: trying to oneshot a feature VS interactive exploration VS boilerplate generation — anuramat, 2025-09
Related videos
- CppCon 2015: Bjarne Stroustrup “Writing Good C++14” · CppCon · 1:40:46
- Change w/ Jim Keller · Tenstorrent · 56:53
- D is for Debacle - The Crucial Story of Vitamin D and Human Health · Ivor Cummins · 1:21:39
- Why Machines That Bend Are Better · Veritasium · 12:51
- "Performance (Really) Matters" with Emery Berger · Association for Computing Machinery (ACM) · 1:04:47
- Analogy as the Core of Cognition · Stanford · 1:08:37
- Robust-first computing: Beyond efficiency · Dave Ackley · 14:46
- The Story Behind Audacity · Tantacrul · 15:20
- Two decades of Git: A conversation with creator Linus Torvalds · GitHub · 41:49
- CppCon 2015: Andrei Alexandrescu “std::allocator...” · CppCon · 1:12:27
- Lexical Scanning in Go - Rob Pike · GoogleFOSSSydney · 51:17
- 🚀 TDD, Where Did It All Go Wrong (Ian Cooper) · DevTernity Conference · 1:03:55
- Bret Victor The Future of Programming · Joey Reid · 32:55
- Transformers, the tech behind LLMs | Deep Learning Chapter 5 · 3Blue1Brown · 27:14
- "Uncle" Bob Martin - "The Future of Programming" · Paul Stringer's Mobile Tech · 1:18:21
- CppCon 2017: Matt Kulukundis “Designing a Fast, Efficient, Cache-friendly Hash Table, Step by Step” · CppCon · 59:33