Even if AI successfully replaces workers, it is doomed to fail. Because if everyone is replaced by AI, who is going to pay for the products/services from the companies anymore? It’s all about being the first one to replace their workers with AI, for short term profit until society collapse.

Reading the comments on the protondb page, a lot of them are using startup options and proton experimental for example. Weird that it has platinum rating if tinkering is necessary though.

The problem with using touch in a car is that you’re supposed to look at the road, not at a screen. Whether something is easily controllable could be a matter of life and death. Something physical you can reach by muscle memory is safer in this context. So yeah the basic controls should be tactile and intuitive. Anyones grandma should be able to figure out how to turn on the wipers if the rain sensor didn’t do its job. Or adjust the fan quickly if the window fogs up.

I think/hope this will change in the future. There’s been a lot of backlash against touch screens, touch buttons, subscriptions for basic functionality, etc. The car industry is struggling right now. At some point I think they will have to make simpler EVs, because normies don’t want ipads on wheels. They want something reliable and familiar that won’t break the bank.

I’m usually playing steam games, and I often will find a solution to make it work on protondb if I have issues. Most of my games I can just install and run though. But I understand it being frustrating if your favourite games don’t work or require lots of tinkering. I have played a few older games outside steam as well. I usually use Bottles for that, as it creates a wine prefix for me that’s set up with DXVK, etc out of the box.

What games? Because a lot of games do work fine, maybe even most of them. The problem is that the outliers are often games that a lot of people are playing (see https://areweanticheatyet.com/). Those games are usually not my cup of tea anyways.

Yeah linux support for ARM SOCs is not ideal. There might be some fork of the kernel working with specific proprietary driver blobs. But in a few years its basically abandonware.

RISC-V is what we should try to make happen as a replacement for x86, instead of yet another proprietary IP.