While open source (well more like “open weights”) cannot easily be enshittified they can absolutely be a stealthily made advertising tool.

For example : if you ask “how do I do X” it could recommand to you based on who paid the company making the LLM more money and you would be none the wiser.

For more obvious bias just ask Chinese models about Taiwan and other problematic subject to the Chinese gov and you’ll see. And the source code isn’t telling us anything about if the LLM weights encode some kind of propaganda, advertising or bias toward/against some entities.

Also I would be interested in knowing what features that you think is missing or should be added in LaManager

Can confirm that I don’t see that, probably on your end

Hey, sorry about that, it seems like I lost internet during the night, it is now fixed.

edit: typo

After putting it to prod on my homeserver and running some test with a slow remote (a friend homeserver). I was able to have ~30 seconds downtime total across all services for a backup that needed 35 minutes to complete.

Usually an offline backups would have needed to have downtime for the entire 35 minutes to ensure a consistant state. Here i have the same guaranties for much cheaper.

note that the situation I have somewhat extreme but it shows well the benefits.

Thanks, don’t hesitate to ask :)