• 0 Posts
  • 13 Comments
Joined 1 year ago
cake
Cake day: May 31st, 2024

help-circle






  • An americano is indeed an espresso (sometimes double shot) with added hot water. Not Italy specific though, it’s just a different thing than filter coffee - but it’s true that in Italy you can usually get the former but not the latter.

    Some other common coffees: lungo (“long”, an espresso that’s extracted for longer so it’s a bit more “watery”), macchiato (“stained”, an espresso with a bit of milk and foam in it, kinda like an espresso version of a cappuccino), macchiatone (“big macchiato”, halfway between macchiato and cappuccino), corretto (“corrected”, with a tiny bit of liquor, usually grappa or sambuca)


  • I’ve been a happy Tidal user for years fwiw. The app is great as is music availability and discovery. I went back to Spotify for a while because I was missing its discovery features like discover weekly, but Tidal has greatly improved since then, and now features a daily discovery playlist (10 tracks, which I greatly prefer to Spotify’s weekly 30), plus 8 custom mixes based on genres you listen to. Track radio is also solid.

    Also, it’s maybe the only subscription service that instead of creating new tiers, merged the two it had before into one, keeping the upper tier’s features at the lower one’s cost



  • Solid advice!

    And remember that “DE hopping” is much easier than distro hopping, as you can install multiple and try them out without reinstalling your system.

    Personally I’m a shill for Plasma, as I think that their motto “simple by default, powerful when needed” is very true. Out of the box, you get a grandma-ready UX that’s pretty intuitive to any Windows or Mac user, but once you start to dig in there’s so many “power user” features. Now every time I’m on a different system I instantly miss all the little QoL that I never even think about, and almost everything is neatly packaged in the system settings or context menus, without having to install extensions or set up a dozen different components




  • In no way given the chaotic context window from all the other models were those tokens the appropriate next ones to pick unless the generating world model predicting those tokens contained a very strange and unique mind within it this was all being filtered through.

    Except for the fact that LLMs can only reliably work if they are made to pick the “wrong” (not the most statistically likely) some of the time - the temperature parameter.

    If the context window is noisy (as in, high-entropy) enough, any kind of “signal” (coherent text) can emerge.

    Also, you know, infinite monkeys.