If it can't run on the metal, is it really an OS that's "native"? Surely an "AI-native" OS runs the AI in ring zero? Is a dockerfile the same as an OS? Always?
These are serious questions, about what is surely not a serious project.
Optimistic of you to assume that future operating systems will run on local compute. Windows 35 requires datacenter level resources, but cortana will sell you minecraft capes with real-time emotionally personalized unskippable advertisements!
That's the whole point of an operating system - to manage local hardware. That's its reason for existing. The moment task switching starts depending on rtt to a cloud provider's servers, it's just a thin client with delusions of grandeur. Fifty milliseconds per system call, and no amount of tflops on the backend is going to save you from that feeling of working through a proxy in 2005
One comment to that demo was: "You cant just end it like this! WHAT DO YOU HAVE BEHIND THE GODDAMN SCENE?!"
Reply was a URL with the exact timestamp to the BEHIND THE SCENES: https://www.youtube.com/live/HG0twQJ7aG4?t=16050. The comment following this reply "was expecting this to be typical reply section clickbait but it's actually legit! Thank you!"
Another comment 'Developer: "Hey check out this funny useless software I made."
Big Tech and Investors: genuinely considering'
I think it's parody? That said all products made with heavy AI use are so inherently bad, and the buffoons so credulous behind their prompts, that it's difficult to suss out genuine efforts by unqualified people vs. people who know it's bad and are doing it to be funny/make that point.
It's kinda like trying to identify actual crypto projects headed by people who have no idea how anything outside cryptographic programming works, versus ones headed by people who know exactly how stupid it is and are just looking to cash a check.
Also the people who hype AI and the ones who backed crypto are basically the same exact circle of bag holders. They'll be out of AI as soon as the bubble pops. I think robotics is gonna be the next one.
To be fair abandoned projects have always been a feature of GitHub and the broader coding community, it’s just now it’s easier to finish the project before abandoning it entirely
There's no practical point to this. System calls and hardware interactions are deterministic. We don't need a probabilistic model to efficiently handle interrupts or write blocks to disk. Ai is great for the user interface layer, but at the kernel level, hardcoded C or Rust will always be faster and more reliable
The only viable scenario for that kind of ji approach would be a profiler that analyzes load patterns and recompiles kernel modules with optimal compiler flags for a specific use case. Dynamically generating kernel code from scratch would kill the system with compilation and ast verification overhead
This gives me that same feeling as 90s-00s Linux .... Sure, your gonna royal mess it all up at some point and have to reinstall (probably sooner than you think)....but look at all the fun stuff you can potentially do with it.
These are serious questions, about what is surely not a serious project.