Comments
By cadamsdotcom 2026-03-0421:021 reply There are two bottlenecks in software.
First is understanding what to build.
Second is getting the details right.
The best software is both the right software, and high quality. It’s bloody hard to do!
So, we do requirements gathering, and try to get the details right (TDD, continuous delivery etc).. it’s surprising to see Gas Town do none of this and optimistically hope agents will converge on good software by just throwing tokens at a wall and hoping it sticks.
So anyway, that’s what they’re ignoring. What are they actually doing???
It’s all for themselves.
Gas Town’s “product” is the warm fuzzies it gives people with money to burn, warm fuzzy feelings of being “at the frontier”. It’s a luxury product for nerds, and the only ones making money or selling anything are the big labs. There’s zero output or benefit to society because that’s simply not the point.
By mieubrisse 2026-03-053:08 I'm working on this! https://github.com/mieubrisse/agenc
I think Gastown is truly special, but I wanted something more focused on learning as I think that's the real bottleneck. So I built AgenC to make it trivial to roll learnings back into your Claude.
By hardwaregeek 2026-03-0417:0410 reply I really want to host a vibe coding competition and see what can actually be made with these systems. Like if we’re doing insane token spends, it better be in service of creating amazing stuff. Can we make an entirely new programming language? Can we make an OS?
By mieubrisse 2026-03-053:25 The guy to watch here is https://github.com/Dicklesworthstone . He's rewritten SQLite in Rust with fixes, written his own Rust async engine with fixes that Tokio doesn't have, generated an insane number of tools for agentic orchestration (indexing of all sessions across all harnesses, on-demand skill storage, agent mail), and is currently building out agent orchestration terminal multiplexer stuff.
Source: been watching both these guys closely, as I've been building my own agent factory focused on security + learning: https://github.com/mieubrisse/agenc
Possible, yes. Easier? I tried to search for YouTube videos of people doing amazing things at blazing speed using Gas Town about a month ago, and couldn't find any. I for sure didn't want to spend hours reading and learning something that I don't know if it even works?
Does anyone have like, projects built using it? I couldn't find "look at the output" types of videos or articles or repos, only "look at the input" types of posts about it.
By antigrav_kids 2026-03-0715:021 reply The 15 year old here used gas town to build a Simon game, but for learning Morse code. It's intended for phones. Still needs documentation touched up a bit.
Scroll the game panel up a bit, turn off the mute button and click start game. A Morse code letter will be played, key it back on the iambic keyer. Pad 1 is for dits. Pad 3 is for dahs. (Dots and dashes for those new to Morse code.) Get it right and the game plays that letter and another random letter back out to you. After successfully sending each sequence back to the game, you get a one letter longer sequence. Just like the handheld Simon game. Link's below
By antigrav_kids 2026-03-0715:06 Forgot, use wired headphones or speaker. Bluetooth delay wreaks havoc adding delays to Morse code sidetones. Like trying to learn piano with delay between tapping key and hearing tune
By steveklabnik 2026-03-0417:261 reply > Can we make an entirely new programming language? Can we make an OS?
I have seen both of these already. I've done the former personally, and I've seen links to at least kernels for the latter.
(I didn't do it via gastown, just regular old "use Claude".)
By hardwaregeek 2026-03-0422:27 That’d also be an interesting data point! What’s the upper limits of vibe coding? Can you vibe code rust? What about an entire programming language toolchain? How about an ecosystem? Can you make a parallel npm?
By rustyboy 2026-03-0417:12 That would be awesome to see! If there was some prize pool like $5,000 to build an operating system through vibe coding tools only and then people could stream themselves on twitch and you could have "sorts desk" type commentators who are collating all that together. I'd watch that and donate a hundred bucks.
By karmakaze 2026-03-0417:33 I would much prefer vibe-coding to be used at the highest layers, not the substrate that we all depend on.
By Sevii 2026-03-0417:41 A lot of it seems to be porting open source projects to rust for other open source projects to consume.
This sounds like a great idea, but I think two questions need to be asked for this to make sense:
I wish we could get to UBI so we don't have to ask those questions.1) Who is organizing/arbitrating? 2) Who is paying for the tokens? Is there some sort of funding or prize?By Etheryte 2026-03-0417:55 In all honesty, if you scoped this well, one of the big players in the LLM space could definitely host a big marketing event on this spin. Get together a bunch of well known industry folks, have them vibe code a working <thing> in a given time constraint, presentations and prizes, lots of marketing.
By xnx 2026-03-0418:18 Crossing my fingers for an open source Chromium browser on Android with extensions enabled.
>Can we make an entirely new programming language?
This is trivial in a few hours with Claude Code
By inerte 2026-03-0418:17 I am designing one, aimed at Claude Code and other AI Coding Agents, and getting the first version lex/parser/compiler was an afternoon project. It was initially a TypeScript toolchain generating TypeScript code.
I keep adding things here and there, a couple hours everyday. Then after about a week I decided to switch the toolchain from TypeScript to Rust, how much work? A 5 minute planning session and a ~20 minutes implementation phase.
Trivial stuff indeed.
By g-b-r 2026-03-052:36 LLMs have a chance to make anything they've been trained on, by outputting it verbatim (or with small modifications).
The question is the degree to which they can produce original things.
By planckscnst 2026-03-051:53 I mostly use LLMs in a zero-touch way - I never actually edit code and I almost never read it. But I do still dive into the details by exploring it with targeted questions through the LLM. Sometimes I go through ridiculously long sessions to get the LLM to "see" the correct/optimal/simplest/etc solution itself. There are many times when it simply never gets there no matter how close I get the horse to the water. I recently did one of these sessions yesterday and it reinforced my impression that systems like gastown and pure Ralph loop style is just not ever going to have the quality I'm looking for, and it's going to cost a lot of money not to get there.
I've honed a relatively decent flow that requires interaction from me for important parts (mostly) while making its own decisions at the not-important parts (mostly). This results in being able to send the agent off on an hours-long dev cycle and have relatively decent results after that need a few minor fixes. I think this is the best style for the current generation of AI
