(_____(_____________(#)~~~~~~

  • 13 Posts
  • 417 Comments
Joined 2 years ago
cake
Cake day: April 11th, 2022

help-circle
  • I have a similar theory. All “AIs” used right now are really just giant matrices that you multiply vectors against. It’s the same concept that’s used in computer graphics all the time, that’s why GPUs are so good for training and running them. To me a more “real” intelligence would need to grow and develop on its own completely organically without any human input and not be just a math problem. It would have to be dynamic and fluid much like a real brain. Neurons would need to function more like individual entities and behave like real neurons rather than just items in an array that get used in simple floating point operations. If you can express any core part of an “AI” as a simple function it’s not really an AI.

    Note that I’m not an expert, I just spent a some years experimenting with different types and combinations of conventional Neural Networks, reading research papers and eventually came to the conclusion that they’re a dead end due to their static nature. This realization actually made me lose interest in “AI” because these things are really just “smart” input mappers that can take well educated “guesses” of what the output might be.



  • I think Micro-Kernels are awesome because the actual Kernel is completely stable and only serves as a tiny (or Micro if you will) layer between the hardware and all the software. All Kernel modules run in userspace as daemons. So if something crashes because of a bug, say the Bluetooth driver, you can just restart that process and continue instead of having a hosed system that needs a hard reset. This is awesome for development because you can just run that buggy driver in a debugger and poke around without having to fear a Kernel panic. Also if you update your System and drivers get updated you don’t need to reboot, you just restart those driver daemons.

    MINIX was really cool. It had a mechanism that could detect dead driver daemons and restart them automatically. They almost achieved full NetBSD compliance and all it really lacked was 64-bit and better hardware support. It technically still lives on because Intel forked it and baked it into every Intel chip. Except they changed the name to “Intel Management Engine”.

    Hurd is GNU’s Micro-Kernel and to my knowledge it’s still being developed but at a sluggish pace. Last I checked they’re still working on full 64-bit support. It is apparently usable enough for Debian and Guix to have a Hurd port though.

    Don’t get me wrong though, I still really like classic Monolithic UNIX-like kernels (especially when they’re simple and easy to compile like NetBSD) but I like Micro-Kernels overall a bit more because kernel modules as daemons is a very interesting concept to me that imo. should be explored more.


  • I might ditch Linux altogether if FreeBSD had better hardware support. There’s only so much I can write and maintain, myself.

    I’ve been feeling that way since they decided to introduce that rust crap in the Kernel. I have a second drive in my main PC with FreeBSD 14.1 and the only things that prevent me from using it as a daily driver are the buggy amdgpu drivers and the terrible bluetooth stack. My dream would be a Micro-Kernel system though. Too bad MINIX was dropped. Maybe Hurd will become usable in 20 years from now lol.



  • Because UNIX-like systems such as GNU/Linux or the BSDs are FOSS (free open source software) where the user owns the means of computing. They’re proven to be extremely flexible while consuming few resources. Not only do they cost nothing, but the user also has access to all the Source code which allows anyone to break and bend the system to fit any shape. For instance the entire Internet right now runs on Computers with some custom variant GNU/Linux or BSD. Every modem/router you’ve come across either runs some version of GNU/Linux or OpenBSD. Every PlayStation starting with 4 runs FreeBSD, Sony just modified it to suit their needs. Netflix’s entire infrastructure runs on computers with FreeBSD. Android and ChromeOS are built on top of Linux, Google just modified it to fit their needs. Valve’s Steam Deck runs SteamOS which is a custom variant of Arch Linux. Etc.

    Literally the only scenario where Windows Server would make sense is if you’re a company that relies on decades old software that only runs on real Windows and most if not all your clients also run Windows. And even then it’s more of a “putting up with it” thing. Legitimate Windows Server licenses cost so much not because it’s quality software, but because Microsoft knows there is a lot of money to be made from people who are locked into their ecosystem. If you use Windows then Microsoft owns your computer and they call the shots. They just let you use it. Not to mention Windows phones home all the time and shows ads in many places and there is nothing you can do about it. If you have the choice why would you want to run a Server for a communist cause on such an OS?

    UNIX-like systems aren’t perfect, but they’re much better than the alternatives.













  • I once showed a person who was constantly scrolling and sitting through ads on their phone that those ads can just be blocked with little effort. They were interested in the idea so I introduced them to Firefox and uBlock Origin and I even helped them setting it up. Days later I meet them again and notice they’re scrolling through ads again. I ask them why they’re using all those apps instead of the Firefox we set up. They told me that using Firefox is difficult because dealing with URLs and setting up bookmarks is complicated. I can’t ever imagine being inconvenienced by a different way of doing things to the point where I go back to watching ads.