Great work, I hope your efforts pay off. But honestly, instead of building an operating system from scratch, I think it might be more sensible to work on improving the management of existing operating systems.
i would not build your OS like terry davis or mention such things. it wont help. trust me.
also, it might be worth suffering a bit with all the errors because operating systems have the power to completely destroy your hardware with bugs.
expect to spend many years on the problems, even when using AI.
Also, if its not unix and not linux or NT inspired, then what is it? Because if you ask AI to make you an OS without specifying the entire architecture etc. it will most likely follow the road populaire OSes have trodden even if built from scratch
What type of OS is it? micro kernel? monolithic? how about drivers or device interactions? power management? what does the userspace and execution environment look like. how do privilege separation and context switches work, etc. etc.
Not trying to discourage, but there's really a lot of considerations to make. hitting all the bugs and designing it carefully will help inform those, and perhaps lead to nice unique features to really distinguish it clearly from other systems
uh i would say making a OS from scratch because on the first place you are using AI and you need to tell and design the architecture yourself because the AI or the llm what you are using is gonna search and build something from its own limited research from the popular ones. and how do you plan to ensure that the code generated by the llm which you are using did not come from a a existing source code.
Great work, I hope your efforts pay off. But honestly, instead of building an operating system from scratch, I think it might be more sensible to work on improving the management of existing operating systems.
How did you ensure code generated by LLM did not com from Linux source code?
i would not build your OS like terry davis or mention such things. it wont help. trust me.
also, it might be worth suffering a bit with all the errors because operating systems have the power to completely destroy your hardware with bugs.
expect to spend many years on the problems, even when using AI.
Also, if its not unix and not linux or NT inspired, then what is it? Because if you ask AI to make you an OS without specifying the entire architecture etc. it will most likely follow the road populaire OSes have trodden even if built from scratch
What type of OS is it? micro kernel? monolithic? how about drivers or device interactions? power management? what does the userspace and execution environment look like. how do privilege separation and context switches work, etc. etc.
Not trying to discourage, but there's really a lot of considerations to make. hitting all the bugs and designing it carefully will help inform those, and perhaps lead to nice unique features to really distinguish it clearly from other systems
uh i would say making a OS from scratch because on the first place you are using AI and you need to tell and design the architecture yourself because the AI or the llm what you are using is gonna search and build something from its own limited research from the popular ones. and how do you plan to ensure that the code generated by the llm which you are using did not come from a a existing source code.