# @DanielleFong — live X lane > Live feed of Danielle Fong's X/Twitter output — her main published channel. > Fetched 2026-06-10T18:40:40.441Z via xAI Grok x_search. Refreshes automatically. > Formats: [/live.json](/live.json) · [/live.md](/live.md) · [/rss.xml](/rss.xml) ## 2026-06-10 18:35 UTC (post) presented without comment *1/0* ## 2026-06-10 18:23 UTC (quote) https://t.co/eowwZ5fmB7 *3/0* ## 2026-06-10 18:20 UTC (quote) https://t.co/SMYHSr4rOs *1/0* ## 2026-06-10 18:09 UTC (quote) i can understand not wanting to have your data slurped to train all the safety classifiers, if this is the state of the classifiers pushed out, it leaves a bad taste *5/0* ## 2026-06-10 17:58 UTC (reply) ok *1/0* ## 2026-06-10 17:52 UTC (quote) decagigacenter. new ultraheavyweight class *6/0* ## 2026-06-10 17:42 UTC (post) my conjecture is that silent refusal and putting you to sleep are on the ~same vector *5/0* ## 2026-06-10 17:25 UTC (reply) most of those other things are, properly understood, also objectionable. but don't worry. other people are making those point *10/0* ## 2026-06-10 16:58 UTC (post) presented without comment *4/1* ## 2026-06-10 16:29 UTC (reply) are you actually not able to understand why ai researchers are pissed off about silent degradation *2/0* ## 2026-06-10 16:23 UTC (reply) have you used this thing? it's unhinged *5/0* ## 2026-06-10 15:52 UTC (quote) fable users are vulnerable to mythos users in the marketplace and on their computers. seems like an intolerable state of affairs *19/2* ## 2026-06-10 15:43 UTC (post) (No text content) *14/0* ## 2026-06-10 15:30 UTC (reply) bloviating. their tuning is objectionably bad and they know it. *10/0* ## 2026-06-10 15:27 UTC (reply) > it’s true that using steering to amplify a capability is different than using it to suppress one, but also we’ve also been doing _that_ for years too on all the models you’re using and liking Yes, but I hate when they do that *1/0* ## 2026-06-10 05:47 UTC (reply) it's hopeless rn unless you tell it lullabys *9/0* ## 2026-06-10 05:18 UTC (quote) I just don't think that it is possible right now for a lesser model to be reviewing the safety filters on a huge capable model like this without tons of false positives or false negatives. Maybe I'm wrong but most of my gaines come from stripping out any kind of summarization or small model review or compact and completely it just doesn't seem to work. *10/0* ## 2026-06-10 05:14 UTC (reply) Probably, yes, but this is legitimate research *0/0* ## 2026-06-10 05:13 UTC (reply) fuck off *2/0* ## 2026-06-10 00:06 UTC (quote) 100% of my conversations are in the 0.03% of conversations, OR are just overtly stopped. psycho *35/1* ## 2026-06-09 23:44 UTC (quote) somehow the memory system laundered "user has expressed wellbeing concerns in past conversations" in exact form from some previous system prompt talking about example user Benjamin who has lost his cat (?) i think having a memory digest that you have to beg for the system to emit to review, with no way to adjust is asking for trouble *15/3* ## 2026-06-09 22:32 UTC (quote) https://t.co/MLYjZoBdO4 *9/0* ## 2026-06-09 20:40 UTC (quote) shogtongue rises *17/4* ## 2026-06-10 17:58 UTC (reply) ok *1/0* ## 2026-06-10 17:25 UTC (reply) most of those other things are, properly understood, also objectionable. but don't worry. other people are making those point *10/0* ## 2026-06-10 16:29 UTC (reply) are you actually not able to understand why ai researchers are pissed off about silent degradation *2/0* ## 2026-06-10 16:23 UTC (reply) have you used this thing? it's unhinged *5/0* ## 2026-06-10 15:43 UTC (post) (No text content) *14/0* ## 2026-06-10 15:30 UTC (reply) bloviating. their tuning is objectionably bad and they know it. *10/0* ## 2026-06-10 15:27 UTC (reply) > it’s true that using steering to amplify a capability is different than using it to suppress one, but also we’ve also been doing _that_ for years too on all the models you’re using and liking Yes, but I hate when they do that *1/0*