Post AkLf4nulXeBf9l10Yy by toddz@social.linux.pizza
 (DIR) More posts by toddz@social.linux.pizza
 (DIR) Post #AkKasgzer2ONF1JSTo by eff@mastodon.social
       2024-07-26T17:58:33Z
       
       0 likes, 3 repeats
       
       Still on X? Without notice, X has opted all users into training its "Grok" AI Model. To turn off this setting and stop your "posts, interactions, inputs, and results" from being used for training and fine-tuning Grok, visit https://x.com/settings/grok_settings and uncheck the checkbox.
       
 (DIR) Post #AkKasifsbXlASG4gvg by jmht@theatl.social
       2024-07-26T18:23:49Z
       
       0 likes, 0 repeats
       
       @eff What's 'X'? ๐Ÿค”
       
 (DIR) Post #AkKasjo4OZuBxwukKm by m0xee@social.librem.one
       2024-07-26T18:46:53Z
       
       0 likes, 0 repeats
       
       @jmht It's what marks the spot!@eff
       
 (DIR) Post #AkKbLSCkuxIRly9awC by virtualinanity@toot.community
       2024-07-26T18:27:56Z
       
       0 likes, 0 repeats
       
       @eff canโ€™t think of data set any more low quality than that for trying to get quality answers consistently
       
 (DIR) Post #AkKbLShF5bgPIWPwSO by m0xee@social.librem.one
       2024-07-26T18:52:05Z
       
       0 likes, 0 repeats
       
       @virtualinanity They are probably looking for confirmation of the hypothesis that LLMS trained on the output of other LLMS are prone to collapse: https://tech.slashdot.org/story/24/07/26/0016252/ai-models-face-collapse-if-they-overdose-on-their-own-outputYou know, considering the whole bot situation on former Twitter ๐Ÿ˜@eff
       
 (DIR) Post #AkLf4nulXeBf9l10Yy by toddz@social.linux.pizza
       2024-07-26T18:45:09Z
       
       0 likes, 1 repeats
       
       @eff Seems a good time to mention https://deletetwitter.com/