Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)UU
Posts
2
Comments
180
Joined
4 mo. ago

  • I actually don't. :) it's the first thing I disable on my phone. I only capitalize "I" specifically, and proper nouns.

    again though, my issue is one with producing the Unicode character of the specific em-dash glyph. maybe some phones automatically convert dashes to em-dashes based on context but I personally wouldn't know how it's done.

  • no no, you misunderstand. I'm not talking about the verbal flourish of dashes for interjection - I use those all the time, in this very sentence - I'm specifically talking about producing the specific Unicode character (—) instead of just using the normal ASCII dash (-). the only way I can make the actual em-dash character is by long-pressing. if I do -- or ---, it's just a sequence of normal dash glyphs for me.

  • it was an art piece, not a serious product idea. they weren't pitching it to Trojan, all it purported to do was make a "statement."

    it's like an OmegaMart product, basically.

  • there used to be platforms that would show livestreamers on a map, so you could get angles on protests as they were happening. I remember using it during the George Floyd protests in 2020. terrible for privacy, but it was incredibly useful to get an unfiltered view from the ground.

    do any of those still exist? or did they shut them all down?

  • this really is a model/engine issue though. the Google Search model is unusably weak because it's designed to run trillions of times per day in milliseconds. even still, endless repetition this egregious usually means mathematical problems happened somewhere, like the SolidGoldMagikarp incident.

    think of it this way: language models are trained to find the most likely completion of text. answers like "you should eat 6-8 spiders per day for a healthy diet" are (superficially) likely - there's a lot of text on the Internet with that pattern. clanging like "a set of knives, a set of knives, ..." isn't likely, mathematically.

    last year there was an incident where ChatGPT went haywire. small numerical errors in the computations would snowball, so after a few coherent sentences the model would start sundowning - clanging and rambling and responding with word salad. the problem in that case was bad cuda kernels. I assume this is something similar, either from bad code or a consequence of whatever evaluation shortcuts they're taking.