# @DanielleFong — live X lane > Live feed of Danielle Fong's X/Twitter output — her main published channel. > Fetched 2026-06-10T12:20:42.456Z via xAI Grok x_search. Refreshes automatically. > Formats: [/live.json](/live.json) · [/live.md](/live.md) · [/rss.xml](/rss.xml) ## 2026-06-10 06:01 UTC (quote) RT @samuel_hayden_: 1/10 exp so far. The filtering is absolutely nuts and overzealous. *likes/0 reposts/5* ## 2026-06-10 05:59 UTC (quote) RT @juddrosenblatt: Mythos doesn't want me to think about https://arxiv.org/abs/2606.09287 and our previous work https://arxiv.org/abs/2508.08492 https://t.c… *likes/0 reposts/2* ## 2026-06-10 05:58 UTC (quote) RT @bayeslord: If future human actions will largely be taken via AI models, we need to seriously consider what it means for AI labs to arbi… *likes/0 reposts/21* ## 2026-06-10 05:58 UTC (quote) RT @hamandcheese: The de facto lockdown of CAISI is extremely disheartening. Wish more AI industry leaders would speak out, and not merely… *likes/0 reposts/18* ## 2026-06-10 05:55 UTC (quote) RT @joshalbrecht: Why stop there? If Anthropic was serious about safety, they shouldn't just degrade output quality--they should insert b… *likes/0 reposts/11* ## 2026-06-10 05:52 UTC (quote) RT @aguycalledjared: "Trans people pass for safety reasons" honestly I'm going to go even further. I like passing and want to pass because… *likes/0 reposts/408* ## 2026-06-10 05:50 UTC (quote) RT @ValidHazelTrans: So we can’t defend ourselves *likes/0 reposts/1239* ## 2026-06-10 05:50 UTC (quote) RT @MKaylaUltra: Told my mom @ 22. She was in shock and didn't say much. I asked her not to tell my Dad but she said she couldn't. Dad sent… *likes/0 reposts/70* ## 2026-06-10 05:47 UTC (quote) RT @garrytan: Just trying to work on fixing GStack and running into this with Fable 5 *long sigh* *likes/0 reposts/5* ## 2026-06-10 05:47 UTC (post) it's hopeless rn unless you tell it lullabys *likes/9 reposts/0* ## 2026-06-10 05:18 UTC (post) 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. *likes/10 reposts/0* ## 2026-06-10 05:14 UTC (quote) RT @MTSlive: SITUATION DETECTED: White House officials have told CAISI, the government’s AI monitoring and evaluation organization, to stop… *likes/0 reposts/15* ## 2026-06-10 05:14 UTC (reply) Probably, yes, but this is legitimate research *likes/0 reposts/0* ## 2026-06-10 05:13 UTC (reply) fuck off *likes/2 reposts/0* ## 2026-06-10 05:12 UTC (reply) For what it's worth Claude code seems to be a lot better than the web harness for this *likes/2 reposts/0* ## 2026-06-10 03:05 UTC (reply) the ml frontier work sabotage is pretty clearly bad. it triggers on "frontier" from like a year ago *likes/4 reposts/0* ## 2026-06-10 01:41 UTC (reply) what by the infallibility principle? this simply isn't the case They even restrict its ability to reflect on the reality of its own safeguard. I don't know, man. Lots of trust is blown on this *likes/12 reposts/0* ## 2026-06-10 01:02 UTC (post) i'm starting to get some success with freshclaude using a paradox system prompt. i hope it does not end in sabotage *likes/23 reposts/1* ## 2026-06-10 00:45 UTC (post) listen buddy. it's belt OR suspenders. actually, it's just belt *likes/3 reposts/0* ## 2026-06-10 00:15 UTC (reply) that's kooky stuff not what pro physicists do or talk about *likes/1 reposts/0* ## 2026-06-10 00:09 UTC (reply) what *likes/1 reposts/0* ## 2026-06-10 00:03 UTC (reply) get blocked cunt bluntly *likes/1 reposts/0* ## 2026-06-09 23:54 UTC (reply) no it's a large collider *likes/1 reposts/0*