Typically LLMs aren't a problem with FOSS with licensing as pretty much anything and everything is free to use, remix, etc.
What is more of a problem is hallucinations, imagining using the wrong rm -rf ~/ command without understanding the consequence, but arguably that's hard to predict. What will always be a problem though, no matter the model, is how much energy was put into it... so that, in fine, it makes the actual documentation and some issues on StackOverflow slightly more accessible because one can do semantic search rather than full text search. Does one really need to run billion parameters models in the cloud on a remote data center for that?
... yet IMHO the real fun comes when you apply YOUR commands to YOUR files.
So yes, please do try in a safe sandbox first then when you want, when you are not rushed by a project start a terminal right there from the comfort of your desktop, then PLAY with your files after doing a backup. Trust me it won't just be fun, it will be truly empowering.
How does SteamOS do on my AMD 7950X3D with my 4090? Oh it doesn’t support those?
Hmmm maybe I'm missing something but doesn't Steam work on these? Doesn't Proton work on these? Doesn't KDE Plasma work too? Basically what's missing for you to consider it working if arguably the most important do work on it?
Asking as I use those 3 daily on my Debian desktop and... it just works. Anyway, back to play Clair Obscur, thanks a lot to Valve.
The command line is precisely trying to address this, providing not isolated apps but commands that are flexible and can be stitched together so that most needs are cover. Think of it like Lego blocks made out of text, that do stuff to your files.
I mean... I don't play Doom due to DRM but also to be fair Expedition 33, which I'm enjoying quite a lot on Linux, is "just" 40ish hours whereas Doom is (check notes) less than 20hrs to complete? WTH. Nope.
Been playing for years on a nearly daily basis, from Baldur's Gate 3 to Elden Ring to now Clair Obscur... to VR with e.g. Half-life : Alyx. That's on my desktop and often also on my Steam Deck. So... yes, Proton is very VERY good for games!
you cannot transfer is your imaginary internet points
Ironically enough, even though "imaginary" this aspect might be key to moderation. Assuming (and that's a flawed assumption) that people would upvote/downvote based not on their opinion but rather on how healthy/unhealthy to the discussion a comment is, then those "points" would be useful to see above/below a threshold one would want to interact, e.g. show content or not (or even now show even as to unfold).
Like I try to highlight, in most cases it's a shitty tool, doing a bad job, trained on stolen data, requiring a TON of energy and often used to put people out of work (and failing at it, cf news above).
So... sure, it's "just" a tool and in theory, it can be made the right way and used in a good context.
It is rarely the case though. Here specifically we are talking about Amazon, a company that has from its inception been built to be a monopoly, relying on AWS a service that is basically destroying the Internet by removing its decentralized nature.
So... again even if the tool would in theory itself be used the right way, build the right way, the company using that tool is problematic.
work remotely and keep my kids entertained while circumnavigating
FWIW depending on your work you can do a lot of that on the cheap, namely if you work is not heavy bandwidth or latency dependent, code and voice do not take much. You can get a lot of resources offline too, e.g. Wikipedia, Stackoverflow, etc in a convenient package with Kiwix. Download this at the port or prior to the legs of the trip where you don't expect to have good connectivity then update at the next point. It's honestly a matter of hours at most. I do it before every trip and it gets easier every time.
My suggestion anyway for kids entertainment is also offline entertainment, e.g. GCompris but even content. Again you can put Wikipedia from Kiwix on your then local WiFi (no Internet, just all devices on the boat) with a small RPi Zero (low energy consumption) with a 1TB microSD card (so cheap now!) but also a media server with all the videos you want from Internet Archive. There is a TON of content. Once there they can watch with any media player that supports network play, e.g VLC or mplayer.
TL;DR: 1TB from the middle of nowhere on the cheap is indeed tricky but 1TB from a good connection THEN go offline is actually both very easy and more than enough to be entertained for months, if not decades with e.g. Gutenberg project!
Typically LLMs aren't a problem with FOSS with licensing as pretty much anything and everything is free to use, remix, etc.
What is more of a problem is hallucinations, imagining using the wrong
rm -rf ~/command without understanding the consequence, but arguably that's hard to predict. What will always be a problem though, no matter the model, is how much energy was put into it... so that, in fine, it makes the actual documentation and some issues on StackOverflow slightly more accessible because one can do semantic search rather than full text search. Does one really need to run billion parameters models in the cloud on a remote data center for that?