Rephrasing a common quote - talk is cheap, that’s why I talk a lot.

  • 0 Posts
  • 985 Comments
Joined 3 years ago
cake
Cake day: July 9th, 2023

help-circle


  • a(n effectively) non-deterministic

    Almost started to type an angry response to that.

    This lady should feel lucky that it only ran amok in her inbox.

    I have done that with less than an LLM. Just a typo in my Mutt configuration, and a few hundred e-mails were deleted which shouldn’t have been. After that I decided that removing spam is best done by first sorting into a separate mailbox and then manual revision. Which is an experience of plenty of people.

    Which just means that if you use an AI agent (and why not, it appears people do want them), then you should perhaps use many dedicated agents only having access each to its own narrow set of available actions.

    It’s more important with things based on fuzzy logic than it is with scripts. But people use Flatpaks and Snaps and AppImages, for isolation among other things, and I have run Skype from separate user under Linux in the olden days (it was such a stupid fashion, everyone wanted Skype, but everyone also considered it proprietary spyware, and nobody thought that an X11 client can spy after the whole display and all keyboard and mouse events anyway ; and that fashion didn’t involve running Skype in Xephyr or Xnest, just from a separate user).

    So the thought is not new. These agents should just be used with clear privilege separation, and some uniform way to declare privileges and interfaces for AI agents, and those interfaces simple enough. One can hope.


  • Separation of server styles, server markup and client styles is definitely something Gemini lacks, not having server styles at all.

    But it’s not as much a problem of browsers as it is of the environment in which information is shared and propagated. While we still connect to websites using a browser, those websites will behave however their owners wish, inflating web standards and requiring complex browsers.

    I was dreaming of something like “hypertext Usenet”, and making descriptions of another system I was interested in trying to make, I am still not even close to that, and I’m not sure I’m still interested, because it appears NOSTR now has much of what I wanted in its standards.

    Basically if you imagine a system for propagating posts addressable by ids and with markup inside, referring to styles and containing hyperlinks by ids to other posts, you can throw away the idea of a website, and still have the hypertext web. That markup can be anything, while the URLs in the links leading to images and such (and other pages) are using those ids or are at least Blossom-compliant.

    I think NOSTR of new protocols is the one most likely to eventually attain such functionality. People here wouldn’t like it, I suppose, because of huge intersection with Bitcoin community and because most clients and client libraries are for the web. But there’s now a C client library, functional enough, and architecturally NOSTR is worlds above the thinking of designers of Lemmy, for example.



  • Since I like the parallel between that and Star Wars holocrons - perhaps there should be separate dedicated devices. Without networking capability, but with ability to add “layers” of the model in the form of memory cards or optical discs as media, containing database files. It helps that hardware to efficiently run an LLM is different than what’s fit for general-purpose applications. People use GPUs for that, but these can be purely dedicated computers of a particular kind.

    The downside is that they’d not be usable for other things. The upside is that you can both have something very convenient and not fear what everyone fears. Display a QR code if you want to copy-paste an answer. Something like that.



  • instead I’m here boning up on the Ferengi Rules of Acquisition.

    I mean, Ferengistan is Europe and in wider sense the West in Farsi, so - pretty logical.

    (Which is why I don’t subscribe to the theory that Ferengi are an antisemitic trope. They are a subversive futurist trope, “seeing ourselves through the eyes of others the same way we often see them”.)

    Everyone likes to see themselves as the heroes of some universe.

    It’s also true for some Soviet science fiction, like things by Strugatsky brothers communicate that deep painful wish for “us” to be that society of scientific workers and doctors, and the barbaric and lost people they visit and help to be “them”, but that’s not how the world is. Even the “approved” Ivan Yefremov with his “Bull’s Hour” shows a space colony which is supposedly a remnant of the “capitalist and imperialist” world, yet surprisingly reminisces USSR, while that team of heroes from heaven that comes trying to fix them doesn’t seem like anything from USSR.












  • Because it’s all trade and balance, so it’s probable that such a window into the world (which will have its own censorship) might be allowed. Probably throttled. Probably allowed and throttled depending on some kind of social rating and individual permissions.

    Unlike Tor, it’s not escaping censorship, it’s one portal (it’s in the name) somehow allowing access to a few select “free speech” (quotes mandatory no matter how you feel about actual free speech) directions.

    USSR had tourist permissions and allowed directions, and friendly socialist countries for which it was easier to get such a permission, and unfriendly capitalist countries for which it was pretty rare and involved state security following you, and so on.

    This might be similar.

    If you don’t see how something can be divided into levels of access for different citizens, then that’s just lack of imagination. They will think of a way.