There are so many levels of LLM involvement and once you make this distinction you actually have to devote some of your limited resources to define and enforce it. So I think it might not be strictly beneficial, but possibly also have downsides.

Where would you draw the line at which point it counts as LLM involvement? Having one run as some sort of spell/syntax checker? Doing some research (and being influenced by it) before hand coding something? Having it create a working prototype and then reimplementing it by hand? If you use code from somewhere else to solve a problem and ai use has not been disclosed for this code, do you automatically have to assume LLM usage was involved or is there plausible deniability?

I think in the end you’d have to use too many resources check it and debate where the line is drawn, when the real question is how you retain human knowledge and engagement within your community.

(all this is just based on what I’ve read, I have no qualifications on this topic)

My understanding is that french and British aircraft carriers differ in the way they launch planes. France uses a catapult, which requires the frame of the aircraft to be reinforced, the UK does not.

The UK also is already part of a separate program called gcap to develop their next gen fighters together with Japan and Italy. This is already well underway and work share has been decided on.

As I’ve already mentioned in this thread, the whole thing about the next gen programs is not just about having the aircraft at the end, but also retaining/building up the domestic industry and know-how.

Sweden might be interested to team up (although theyd probably prefer a lighter/cheaper aircraft than Germany), but my understanding is that Germany and Sweden both have similar gaps in their know-how. I think it has to do with some hot parts of the engine, which in Europe France or the UK would be the experts in. You can ofc develop them yourself from scratch, but that is costly and more importantly likely takes too long.

But the military requirements are only half the issue, the other half is building up and retaining institutional knowledge in building auch planes. Otherwise Germany could probably solve their problem easily by just buying whatever gcap ends up producing.

What I find quite fascinating is that on one hand 100 billion is a fuck ton of money, on the other hand it’s like 2 quarters of profit for Google or less than what Meta is planning to spend on data centers in 2026 (they also spend like 70billion on the Meta verse). I guess money wasn’t the direct issue right now, but France (a large nation) will have trouble with funding when they decide to do it by themselves

I think with the German TÜV numbers it is not quite as clear cut.

My understanding is that telsa unlike other manufacturers doesn’t require regular inspections to keep warranties. So it might just be that other brands have just as many faults, but they get caught and fixed during those regular inspections, rather than at the official TÜV. Which wouldn’t show up in that dataset.

I haven’t seen any studies accounting for that discrepancy, but I’d be curious to see whether the higher failure rate persists after accounting for that.

I also tried TrueNAS scale once (but I think at the time it was quite new so maybe not how it is now) and didn’t like it either so I won’t speak about that.

If it were literally just for network storage and nothing else I guess the way to go might just be to hook up a usb drive to ones router (if it supports that). So most people probably do want slightly more or at least the option to expand capabilities without having to fundamentally change their setup.

But in general I would say that yes, the web interface is definitely part of the appeal. I think what something like Openmediavault offers is that someone else chooses sensible presets, it gives an easy to use gui that allows oneself to get stuff done without any major technical knowledge, and it also creates a community with a shared similar setup that can offer support.

I guess it depends on what one wants from the server. Of course you can have a plain Debian or whatever installation and do everything yourself. But for most, especially the less technical knowledgeable, it is nice to have some interface for setup and maintenance stuff.

I (very much an amateur) briefly tried TueNAS scale in the past and didn’t like how they did apps. So I switched to Openmediavault, which since then has served me very well.

With a plugin I could easily add my zfs raid and I use their build in docker compose gui to run the few programs I need.

I didn’t try out others, but there are more options. CasaOS and yunohost already got mentioned, there is also Cosmos or just running a basic server with e.g Debian and maybe adding Cockpit for some management gui.