LoqiLarge language model traffic refers to web requests made by LLM operators to ingest data for use in training AI models, or to answer user queries in "search" mode (i.e https://indieweb.org/large_language_model_traffic
madude, thegreekgeek and ensoyote joined the channel
[artlung]2^ there is no writing edition coming up tomorrow. Messages in any of the chat channels beginning with “[edit]” are typically glitches from Discord.
capjamesgHWC Europe will be on a break until the first or second week in January. Stay tuned on https://events.indieweb.org for updates on when our next event will be!
trwnhi would make the same claim more generally for any microsyntax. i know chris messina is quite proud of being known as "the inventor of the hashtag" (and in my experience gets a little prickly when people suggest things about hashtags), but the same problem exists for hashtags as well.
trwnh@mentions too, have the same problem. the problem being more generally: you cannot extract meaning unambiguously unless you know which conventions are being used and in what context.
trwnhre: "visual noise" i would say that people are doing the best they can using communication formats that are too limited. i'm doing so in this very irc message by saying "re:" to indicate that i am replying or responding to a specific thing.
trwnhmost formats have a sort of separation between header and body, metadata and content. twitterlikes don't let you make arbitrary claims, and you can only control the plain text content. so really it's just one more matroska doll you have to unwrap -- you get an http message, unwrap it for the body content which is usually html, then unwrap the html to get the real content (say a .h-entry .p-content), but you can unwrap that plain text to get
trwnhif the author only has control over the .p-content and cannot set their own html tags or http headers, then they will reinvent equivalents within the plain-text content.
[tantek]@-mentions also had tons of prior plain text usage, blog comments, Flickr photo comments, PHPBB forums etc. many people simultaneously used @-username without anyone top-down declaring it to be a "thing". there was already widespread human linguistic usage in the wild before @-mentions were "formalized".
[tantek]there's a big misconception by a lot of developers / technologists that if only they make the "right" proposal, with the right "pitch" that it will be magically adopted. that methodology is nearly always false — has a horrible track record.