Live data from Hacker News

Parallel AI agents are a game changer

morningcoffee.io

71–80 of 88 posts

Re: Parallel AI agents are a game changer

#71
post #63

Earlier quoted context omitted.

Having 20 PRs open at once doesn't necessarily mean managing 20 agents simultaneously. It can mean, for example, that 2 agents worked for some time in a list of 20 TODO features and produced 20 PRs to be reviewed. They could have worked overnight even. You're seemingly judging from the least generous interpretation, which is not constructive and is also against HN guidelines fyi.

I’m not ok with someone self promoting here at the cost of thousands of people thinking they’re either not smart enough, or are doing something incorrectly. We saw this same pattern during the dot com boom a quarter century ago with self promoters creating a “you just don’t get it” culture which eventually collapsed like a house of cards. What we share should be reproducible by others and we should avoid hand wavey e…

Technology evolves. At some point there's going to be things that other people are doing that you can't replicate yet. That doesn't mean you're not smart enough. But it might mean that you are doing something wrong. Often though, you just got to try different things or wait for methodologies to consolidate and become mainstream.

Even if parallel agents is not something easily done currently, debating about ways to do it is constructive enough for me.

Re: Parallel AI agents are a game changer

#72
My experience with parallel agents is that the bottleneck is not how fast we can produce code but the speed at which we can review it and context switch. Realistically, I don’t think most people have the mental capacity to supervise more than one simultaneous task of any real complexity.

Re: Parallel AI agents are a game changer

#73
post #9

I use Claude every day. When I give it a program that does something straightforward in one file and it writes it from scratch, it does great. When I have it fix issues or add functionality to small to medium sized apps that have a mostly simple design, it does great. When I point it at codebases that are 20 years old and have a lot of indirection in their design due to years of hard lessons learned and a lot (like a…

one thing that i find works really well is to ask it to research things in the codebase and write a plan first. codex with gpt-5 is exceedingly good at doing this. then ask it to write a plan for what it would do with that information, i.e., i want you to research codebase for . then write a plan for how you would achieve given what you have learned.

Claude writes out plans and all that, it's good about that.

Sure would be great if ai agents could learn from conversations. That would really makes things better. I tell Claude to capture things in the Claude.md file, but I have to manually tend to that quite a lot.

Re: Parallel AI agents are a game changer

#74
I totally disagree about the monorepo argument. Mono-repos exist as a reaction to poor modularization and tight coupling which creates a need to update dependencies frequently.

The reason for having dependencies in the same repo as the trunk of the project code is precisely because the dependencies aren't sufficiently generic, too dependent on the project's business domain and so they require constant maintenance.

This tight coupling means that the agent requires more context to solve problems and implement simple features. The need for more context is a problem for agents, not a benefit. Agents benefit from modularization, loose coupling and well-chosen abstractions. These attributes do not correspond to the kinds of complex, tightly integrated codebases which benefit from having a monorepo structure.

Dependencies should be like tools. If you think of a hammer, you can do a lot of different jobs with the same hammer... You can debate whether or not a hammer is the right tool for any given job, but for those jobs where a hammer is the right tool, how often do you need to tweak the hammer itself? A hammer solves a very specific problem but that problem can be generalized to countless different use cases. A hammer would make a good module.

Now if you did a project for a candle factory and let the project business domain leak into the design of your tools/modules; you may build a hammer out of wax to straighten out candles... Then in your next project building a house you will find that this hammer doesn't work for that case and needs to be modified. This is a failure of separation of concerns. The tool was not originally optimized for the specific task of applying blunt force to a limited area; it couldn't do that narrow job very well and that's why it needs to be changed. Had you built a hammer out of steel, it would likely have solved both problems even through it's a completely different use case.

Re: Parallel AI agents are a game changer

#75
post #17

Can this guy, or someone else post a full days (4-8 hours, or what ever is spent in the weeds) stream of work to youtube or something. I just want to watch the process to see what I'm missing. Or if there is anyone that already does that can they recommend it to me. I would appreciate it.

https://youtu.be/xAKVi_jvvg4 Two hours of Web Dev Cody.

Got through about 45 min at 2x speed / some skipping ahead out of pure fascination. Man that's something else. It's like bug-driven-development. Get the LLM to churn out a huge chunk of text, then skim the code for about 10 seconds and say it looks good. Then spend a while testing and hitting one error after the next until it finally seems to work. Repeat.

Re: Parallel AI agents are a game changer

#76

Look, I like AI coding but we're already way past the need for parallelism. LLMs write so much code in such a short time that the bottleneck is already the human having to review, correct, rewrite. Parallel agents working on different parts of the application just compound this problem worse, it's impossible to catch up. The only far fetched use case I can see is swarming hundreds of solutions against a properly desi…

Exactly. With other models that are not Claude, the code generation for an issue takes a minute at most, whereas writing the detailed specification for it as a human takes me days or longer. Parallel code generation is as relevant to me as having a fast car stuck in traffic at a red light.

Re: Parallel AI agents are a game changer

#77

Look, I like AI coding but we're already way past the need for parallelism. LLMs write so much code in such a short time that the bottleneck is already the human having to review, correct, rewrite. Parallel agents working on different parts of the application just compound this problem worse, it's impossible to catch up. The only far fetched use case I can see is swarming hundreds of solutions against a properly desi…

It really depends on the project. For example, there's a lot of thorny devops debugging where I can just let Claude spin for 30 minutes and it'll solve the problem (or fail) with a relatively short final answer. The sweet spot for me tends to be running one of these slower projects on a worktree in the background, and one more active coding project.

Yeah sure, I mean, there always be problems you can swarm..

Re: Parallel AI agents are a game changer

#78
post #17

Can this guy, or someone else post a full days (4-8 hours, or what ever is spent in the weeds) stream of work to youtube or something. I just want to watch the process to see what I'm missing. Or if there is anyone that already does that can they recommend it to me. I would appreciate it.

https://youtu.be/xAKVi_jvvg4 Two hours of Web Dev Cody.

Wow, I didnt expect that dystopias can be so boring.

If somebody like that producing code of like that low quality worked with me, I can see myself spilling coffee or acid on them or their laptop.

Re: Parallel AI agents are a game changer

#79
Counteropinion: neither parallel nor single threaded agents are a gamechanger.

I mean they might be changing the game into prodicing more hard to maintain software, faster, but if that is the game you are playing, I dont wanna participate.

Re: Parallel AI agents are a game changer

#80
post #42

Is it me or the post sounding ( showing! ) that they haven't tried the mentioned approach in real life. Because in real life, one agent tries to fix build issue with rm -rf node_modules and the other is already running a server (ie: npm server), conflicting with each other nearly all the time! . (even if it's not a destructive action, the second npm server will most likely to fail due to port-allocation conflicts!) M…

git workspaces?

u mean worktrees
Post reply on HN