What we’re currently calling AI isn’t AI but just a language processing system that takes its best guess at a response from it’s database of information they pilfered from the internet like a more sophisticated Google.

It can’t really think for itself and it’s answers can be completely wrong. There’s nothing intelligent about it.

What advantage do you get using one PC in this way instead of two?

The GPL also enforces that you need to share your code, which isn’t necessarily what these devs want.

You make some good points but yeah, if you licence something under a license that allows corporations to do this don’t be surprised if they do.

I don’t know if there’s some license out there that allows free sharing of code with a limitation around using it in for-profit products and profit sharing for them and whether such a license would even work.

1. All of the basics should just work well out of the box with minimal tweaking. Yes even NVIDIA stuff.
2. The software center needs a massive overhaul. It feels like an afterthought by people who would rather use a command line.

Given the stranglehold that Chrome-based browsers have on the market, I’d say it’s the opposite.