gtadesktop1

As a developer building a custom 64-bit OS completely from scratch on bare-metal C and assembly, I feel your pain to 100%. Prompting an AI to write a kiddy "Word clone" or boilerplate web apps is the ultimate passion killer. It makes software engineering feel like monitoring a factory assembly line. The only way I found to escape this AI-slop burnout is shifting your focus entirely to low-level or complex domains where AI models constantly halluzinate and fail. When you orchestrate raw hardware registers, write custom USB-HID or xhci stacks and handle direct MMU paging without any framework safety net, AI stops being a driver and becomes a basic typing assistant. You can't vibe-code a stable OS kernel 'cause if you don't understand every single byte in the RAM, the CPU will just triple-fault and instant-reboot. Don't let corporations turn you iinto a prompt manager. Switch to systems programming, compiler dev, or bare-metal embedded stuff. Go back to where you have to understand the machine, not just nudge a model. That's where the real joy of building lives. If you need such a project, try to write a bios-like system for microcontroller like the esp32 'cause there you have to save resources for developers who want to write an os for your bios-like embedded system. No AI just you and the hardware.

Losing access to Claude Code - just imagine it. That's really the main reason people push for local open-source models. If you build everything around a closed cloud API, you're basicly staking your whole career on someone else's rules. Anthropic can change prices or shut down your account anytime, and suddenly, your productivity takes a nosedive. There's only one way out: leran to run your own setup. Right now, getting Claude Code to work well alongside Ollama or Qwen is tough. Context management and system instructions are tricky. But honestly, it's our only real safety net. If you can fire up quantized Qwen 2.5 coder model on your own machine, you control your workflow. Sure, it might not be as fast as Sonnet, but there's no kill-switch, no subscriptions, and no strings attached. Think of AI tools like servers: if you can't self-host a backup, you don't really own your workflow. I worked with a friend on a local AI-Agent that can code, that can talk with you about anything and can control your whole computer, even spotify it can easily search on spotify or you can say "open zapzap" or "open whatsapp" it looks for apps on your computer if nothing found with the name it opens the browser and google if you say "open youtube" it opens the browser with https://www.youtube.com. of course you can say it with your voice what you want, so you haven't write it to it per keyboard if your busy. if you wanna try just write a response here and I talk to my friend if I can make it open source. Sorry for self-promoten here but it's something about the topic. if you have a problem with the self-promoten just write a reply and I'll remove this comment.

Yes true point. But why use Java? I mean you said another language that is an alternative for Java.

.

Starting a startup is a great idea. However, your concerns are certainly valid. But investors will invest as long as what you’re doing is worth investing in—they generally don’t care whether you use AI or not; it’s more about the quality of your work and how much money investors can reasonably expect to make from your startup. It’s like borrowing money: investors see the value of your startup and invest with the intention of either getting their money back in full after a certain period at an agreed-upon interest rate—so-called debt securities—or acquiring shares in your startup that give them voting rights in your decisions and a certain percentage of your revenue. Either way, as soon as your startup gains value, larger investors will come on board. The value of your startup depends on how many securities you issue or on your revenue. Bottom line: Investors will generally come as long as you stick to not using AI and your work is of high quality—which I have no doubt about based on your description. Good luck with your idea.

Yes, that's 100% correct. Security by obscurity is really bad, especially because with open-source projects, hundreds of people can review the code, find bugs, and improve it. According to statistics, significantly more cyberattacks succeed on closed platforms where the code isn't accessible. For anyone who wants to see the statistics, here's a link to some data—or rather, a statement from a security economist: https://www.researchgate.net/publication/220891308_Security_...

yeah that's a true point. But that also means that a human can't code it either. except the person comes from that time so the person has to be over 40 years old and can code or can prompt in the "right" way that the ai can reproduce this.

Vibe coding sounds easy 'cause the ai makes all the work. but it's much easier to code per hand because ai can make mistakes that a human don't see. also you have to prompt in the right way. if you don't prompt "right" the ai makes a thing you don't want you can write the ai "write me a word clone" and the ai gives you a simple texteditor code where you can write text and save it but you haven't hypermarker, ruler, image inserts, video inserts, link previews, etc. so you have to describe the ai exactly what you want into the least detail. thats very complex so most developers don't vibe code or vibe code only easy parts or complex parts that takes usually months to develope but a ai can do this in 10 seconds if you prompt "right". usually you use prebuild projects.

I'm working on a own x86_64 UEFI Operating System (OS). The idea is to make an OS that looks like Windows Vista but behave like Linux, so it's a mix of Windows and Linux. I've startet three weeks ago with a custom bootloader and the legacy bios VGA textbuffer and now I'm working on a real USB-HID (USB-Human Interface Device) and have troubles with the keyboard. I also have problems with booting from real hardware. If anyone is good at low level developing the project is source available and I appreciate collaborations with other os developers. Link to Codeberg-repo: https://codeberg.org/vntx-labs/VeloOS

Hope that is no problem that I use this thread a bit for self-promotion. If so just write a reply I'll remove this comment if this is a problem

Fair point, my bad! I spent the last few days writing a custom compiler for my own OS project and the OS itself and got so deep into the "talking to the machine" mindset that my grammar slipped. I definitely meant "it". Thanks for catching that.

I used Claude too, but I was banned so I searched for new like you yet and found Google AI Studio. It has a Playground-Mode for the newest models you usually have to pay but in the Playground-Mode you don't pay one cent. You can use Gemini 3.8 Flash and you can give him system instructions. For example you can say "You're a AI Agent that helps me with my mobile app" and he helps you with your mobile app. It has a context window up to 1 million tokens and it's fully free. If the limit is reached you can only click on the "+" icon on the top right and say him what you did attach all project files and tell him what you need. He reads the files you attached and helps you like the old chat.