I know an argument can be made about the brains just being devices that respond to stimuli, but brains literally have parts to process emotion. Most LLMs cannot change themselves, therefore they can't learn. The way to simulate learning is to provide a tool to store memories.

The difference is that the target has no capabilities to understand, it only knows how to respond. It responds so good that people believes that it could be capable of more than responding. It's a large matrix with weights. It has no other facilities besides responding. Most aren't even changed.

I have two MD players and I’m so sad my stereo MD player has trouble focusing

Did stuff for a robotics team and we just had people using GitHub Desktop instead of teaching them git commands. I could see people getting by easily without touching the Git CLI

Gonna be honest, it feels like there will still be a lot of AI still used instead of paying people. There are already models out there where you can generate images locally on modest hardware. Either that or pay people on Fiverr to generate images for very little money.

By using any number of Git UIs that handle the commands for you

Beneath it all, JavaScript is a good and light language. JavaScript has just been used for things it wasn’t designed for and that causes pain points in the language

I got blocked lol

I know semantics got confusing, but you agree that:
A. Most compositors share code and/or shared libraries (via the reference implementation of the Wayland protocol)
B. Because the reference implementation is used by these compositors, the compositors would share dependencies from libwayland.so?

So in short, libwayland.so is what does the communication (if you use the reference implementation). Whether it's libwayland-client.so or libwayland-server.so. You can create a substitute for this library that communicates with the Wayland protocol, but most don't. Ergo, it's shared.

I totally agree. Wayland is just a protocol. I said it was exclusive to Linux and FreeBSD since they both have mature implementations.

libwayland.so is the reference implementation of the Wayland protocol. wlroots is a library to assist in developing a Wayland-compliant compositor. While compositors can reimplement the Wayland protocol, most use the reference implementation. If they use libwayland.so they are using a shared library

Okay, so I think we might be getting a little bit confused. A majority of compositors look to be dependant on libwayland-server. libwayland.so is a shared library. It implements the Wayland protocol. While you can just not use it, most compositors look to use it.

...but that means that they would share code and/or shared libraries. Yes, there are independent implementations. It looks like the majority of compositors have a dependency on libwayland-server which would be both shared code and a shared library.

Just looking at Alpine Linux, wlroots is used by cage, cagebreak, labwc, mangowc, river-classic, sway, and wayback. Wlroots depends on libwayland-client and libwayland-server. Iirc Xfce is using wlroots for its Wayland session and Budgie is using wlroots for Magpie, their Wayland compositor.