

I can’t imagine self hosting an LLM-based search engine would be too viable. The hardware demands, even for a relatively small quantised model, are considerable. Doubly so if you don’t have a GPU to accelerate with.


I can’t imagine self hosting an LLM-based search engine would be too viable. The hardware demands, even for a relatively small quantised model, are considerable. Doubly so if you don’t have a GPU to accelerate with.
Not just, but he literally advertised himself as not being technical. That seems to be just asking for an open season.


Ease of installation/use, I think, is the main big one, and one of the biggest obstacles.
People who want to give self-hosting a try aren’t going to be particularly fond of having to jump through a whole bunch of different configs, and manually set everything up.
They want something that they can just set up and go, without having to deal with server hosting, services, and all of that. Something you can just run on your computer, leave it be, and use it with relatively little fuss.
Second to that, would definitely be a case of better documentation/screenshots. A lot of self-hosted things, like Lemmy, didn’t provide much documentation of what the actual user side of it does, only what you need to do to set it up, which isn’t going to make me want to use the software, if I have no idea what it’s supposed to do, and how it compares to other things that do the same.
Or it turns out you accidentally left caps lock on, and now you’re locked out for a few minutes.
That’s why you get the best of both worlds, and combine them to use tabspaces. Everyone will hate you equally, achieving true equality in the process.
Some time ago, never mind how long precisely, Plex were trying to legitimise themselves, by adding streaming from official sources, etc.
I would be curious if this is meant to be a deterrent, or just to look like one by making piracy expensive, so they can eat their cake and have it too.