AbhiramaVS

The thing is, stuff like projects with a lot of low level code (such as building your own programming language) is not THAT good at being AI generated. Personally, I use the like manual version of claude where you have to like approve everything it does because when I was testing it out myself (for your "real world" data), the LLM would write code in a HORRIBLE way that would be redicioulously difficult for a human to optimize in the future because, there is a limit to how fast the code that an LLM generates can be. often times what ive seen is that it just does things the "quickest" way or the most "technically correct" ways (i.e functional programming or using a lot of pre built APIs which introduce overhead etc) which arent the best ways or the most optimal ways.

An experienced human can write code better than an LLM for low level tasks almost 100% of the time, the only problem is its REALLY slow for a singular person to write code when compared to like an LLM running at 600 tokens per second or something. The best combination of the two is (in my opinion) running claude on manual mode cus if it does something really bad you can either fix it yourself or reprompt it, but there is a LOT of active things you have to be doing.

>Nope. haven't done it. But I have done things more complex then language design. And I can't show you because my employer owns it. But FYI language implementation follows very common patterns and this makes it very very easy for LLMS to create one. A beta of a language can be done in about a week.

its not about the fact that the thing you are doing is "more complex" (in my humble opinion the only thing more complex than language design is building an OS) but its more of the fact that if the LLM writes one line of code that is not like optimal for performence, your language will like leak memory or it will run really slowly, or something will happen suboptimally. you cant just leave it autonomously for large periods of time (I have tried and regreted this) but you can still use it, as long as you read over and esnure that the code is optimal. if you dont do that we are just moving backwards in time and the code you write is very unmaintanable, slow, leaks memory or is volatile, and impossible to scale up or down.

Ah yeah that makes more sense. Will look into this

Again I am a bit confused. Its not like the agent is driving the human, you are driving the agent. If your agent has to come back to you with periodic updates, then you are fundamentally not using it correctly. You should be actively monitoring the code produced by the agent, as frequently it is not doing stuff in an optimal way.

Additionally (i have no idea if this is intentional or not), the order of my words in the readme seems to be flipped here.

This line > Additionally, around 80% of the raw code in this repository was produced with agentic coding tools

came prior to this line > I would not consider myself a "vibecoder", or jaithon as "ai slop"

Phrasing it the other way like you just did makes it seem like I contradicted my self heavily, however in reality I had the second line with a bunch of reasoning in the readme.

I can speak for Jaithon but this is standard across most languages who have like immutable and mutable variables.

There is def nothing stopping you from using var everywhere, but (let) lets the compiler like catch any mistakes you did not mean to do. In jaithon, let is whats known as shallow: it prevents (in ur example) names = ... but the refrenced list is still mutable. Currently jaithon doesnt have anything like a read-only parameter so just knowing list[str] doesnt tell u anything about if a function mutates the list.

what do you mean? I didnt understand your question

imo you should use these tools to the extent that you still understand the code. i put this on my readme, but how I use these tools is I first draft the code for wtv Im doing manually (like a bug fix, new fewture etc) with a ton of comments so the LLM can understand what I ment to do, and then I just let it finish up the code with implementation, integration with rest of codebase, optimization if the code I wrote was terribly written and like also just to tell me if the bug fix is wrong etc.

i think this is an appropriate use of AI but would love to hear other opinions

yup!

I recomend you take a shot and try. Originally, I built Jaithon pre chatgpt (2022-2023) and in doing so I read a couple of books my dad owned about language automta and architecture. I stopped working on jaithon not because I couldnt code or anything, but more so because it took so long to actually push something that I would rather just work on a different project. idk maybe im just ADHD but thats just how it went for me.

Anyways, back to your question, the architecture of jaithon is probably the most complex thing, and AI (atleast now) cant do this reliably. Syntax as well is something that is very abstract that AI isnt good at choosing. there is a lot more behind the scenes than simply oneshoting a prompt with opus and expecting it to work (spoiler it wont work well at all.)

Would love to see you try to do this with pure AI

free to believe whatever you want.

Yes! I agree. I wish github made this a standard or something like that, I personally just did the best I could with disclosure. Thanks for the kind words!

AI Slop is more of a convention than an actual definition. There is no standard definition (from what I know) that is widly accepted for the definition of AI slop: some people might say any use of AI is slop or others might have a more liberal opinion (like me) where if AI is properly used and generates good results, it is not considered slop.

In my personal opinion (you are 100% free to disagree as this is just an opinion and not fact) the 'hard' part of programing is coming up with the proper architecture, and ensuring the codebase is maintainable and is able to be built on top of. Additionally, the word slop in my opinion signals that the product is terrible. In my opinion, the product of Jaithon is not terrible (like some AI generated code is) and as I have not only been highly transparent with the use of AI and also used it with an appropriate maner with limits, by my definitions of it, this is not considered slop.

Additionally, the vast majority of the CODE was carried out by AI, not the implementation. Architecture & language automota were actually created in 2022-2023 while I was in 8th grade. AI was only used rather recently to speedrun development.

Hope this serves as further clarification by what my readme and my words ment.

Unfortunately I did not do this :( I think a couple people pointed out but .jai already exists on github (I did not know about this when I first made it lol) and for the readme and stuff I did it via python for the syntax highlighitng lol.

Maybe one day I can plead with the github gods to accommodate for jaithon :)

More like rust than both of those tbh. When i first made jaithon i named it as such cus back then for one the syntax was vastly different and for two I only knew python and java lol. more recently I changed the syntax to be more similar to rust and js etc.

Jaithon wasn't necessarily based on a particular language or multiple languages even per say, its just my preferences from all the languages ive coded in the past :)

Thank you so much!! That was completley my intention and the reason why I was building it, and im really glad you liked it! To be quite frank, the syntax is more similar to rust compared to python in a couple of ways, but python is what inspired me to build the language.

The syntax is just a combination of my personal preferences :) seems like we share a bunch in common!