Posts by buzz@andersen.social
(DIR) Post #561559 by buzz@andersen.social
2018-08-30T02:05:42Z
0 likes, 0 repeats
“Which is of course why it is so dangerous to join a mob… any mob, no matter how “noble” the stated purpose may be. The mob is almost surely led by a malignant narcissist. And that narcissist is using you.” That quote comes to mind for me a lot when thinking about social media these days.https://yesivebeenthere2.wordpress.com/2014/03/24/narcissists-sexists-racists-child-abusers-internet-trolls/
(DIR) Post #AIycL6hHFFc7oL4RWa by buzz@andersen.social
2022-04-30T04:00:56Z
1 likes, 0 repeats
@shadowfacts I remember the same article! I believe it was this 1996 Fast Company piece. https://www.fastcompany.com/28121/they-write-right-stuff
(DIR) Post #APL9T5jMnWzCj6bfcG by buzz@andersen.social
2022-11-06T19:16:16Z
0 likes, 0 repeats
Little Reichstag Fires Everywhere https://gizmodo.com/donald-trump-homeland-security-report-antifa-portland-1849718673
(DIR) Post #APT36hyAvZcXTdxIdE by buzz@andersen.social
2022-11-10T16:29:09Z
0 likes, 0 repeats
Reading through stuff about Mastodon’s performance issues and pretty much stopped when I got to “Mastodon is a Ruby on Rails application.”Rustadon, anyone?
(DIR) Post #APT36nNsnQ9gGp1Otc by buzz@andersen.social
2022-11-10T16:30:40Z
0 likes, 0 repeats
This really is like 2007 again. I fully expect someone to come out of the woodwork now and take umbrage at my Ruby bashing.
(DIR) Post #APT36q2MvrjEUxylHM by buzz@andersen.social
2022-11-10T16:31:55Z
0 likes, 0 repeats
Oh, LOL, it actually exists! https://github.com/rustodon/rustodon
(DIR) Post #AT9mPvJzvYDDlcoKvo by buzz@andersen.social
2023-02-28T23:26:11Z
0 likes, 0 repeats
God, this is several levels of dumb. https://www.theinformation.com/articles/fighting-woke-ai-musk-recruits-team-to-develop-openai-rival
(DIR) Post #AT9mPw3NCqVa2M2Zeq by buzz@andersen.social
2023-02-28T23:30:08Z
0 likes, 1 repeats
Level 1: The idea that any form of content moderation is “wokeness”Level 2: The whole “wokeness” canard in general Level 3: The implication that these language models are anywhere near sentience, and thus capable of “wokeness”
(DIR) Post #AT9mQ0dg9nvUG73cno by buzz@andersen.social
2023-02-28T23:31:56Z
0 likes, 0 repeats
Meet my new language model, trained exclusively on cancelled comedians who aren’t afraid to tell it like it is.
(DIR) Post #AVKsa93gWQEi3rd3gG by buzz@andersen.social
2023-05-05T05:56:55Z
0 likes, 0 repeats
@gruber Oh, and: it supports freaking quote tweeting. Which, I would argue, is central to how some very engaged users use Twitter. I really think the Mastodon community made a big mistake in assuming quote tweeting was only associated with abuse.
(DIR) Post #AVKss9NS4j9iGLSbQm by buzz@andersen.social
2023-05-05T06:06:20Z
0 likes, 0 repeats
@anderspuck @gruber Yeah, that’s what I keep hearing. But they’re already behind the curve on this and losing mindshare.
(DIR) Post #AW66yYj9d5ZpDvPczw by buzz@andersen.social
2023-05-27T16:47:12Z
0 likes, 1 repeats
It’s almost as if these large language models are extremely unsuited to the primary application being imagined for them: knowledge retrieval.
(DIR) Post #AW66yb2j1IQ6PCaHmS by buzz@andersen.social
2023-05-27T16:52:52Z
0 likes, 0 repeats
The choice of the term “hallucination” to describe what is happening with these language models is fundamentally misleading, and feels calculated to obscure their probabilistic nature.Characterizing the fabricated output as “hallucination” means, by implication, that correct output is “knowledge”—that the incorrect results are aberrations, not the result of the exact same blind processes that sometimes produce factually sound results.