fama

@creatine_cycle He'd be best friends with the palantier sword guy who said he's geographically monogamous

if genghis khan were alive today he would have been a thiel fellow

@pleometric A purple penguin, or so I've heard

I have watched a few people go insane because they went giga viral, I wonder what triggers that?

@ToughSf I thought of all the trouble we go through to even collect a gram atm so 10 tons is a loooot

The first crewed interstellar starship ready to go, with a pristine debris shield and tanks full of antimatter. 10 tons of antimatter is enough to set an entire hemisphere on fire from orbit so it's dangerously close to Earth here! #space #art by Louis_Rtr blenderartists.org/t/interstellar…

@mattparlmer Its by design (wink)

I don’t think the LLM-caused delusions are gonna get any better, people with poor cogsec seem to be digging ever deeper holes for themselves

I wonder if @nearcyan finally figured out the cure to how to sleep without swallowing a million pills. Bro its midnight and I can't sleep

@dwarkesh_sp Ask him how many jackets he owns?

Tomorrow

@RuiHuang_art Man I love your art

Will you watch 30 min of this?

#JoJosBizarreAdventure #SteelBallRun

@SHL0MS I can't tell whether you're being serious or not. This app is cooked

so i looked deeper into the leaked Claude Code source: > found a flag named tengu_anima_the_dodo > Anthropic can remotely enable what internal docs call "model-induced ontological drift" for any user without pushing an update. flip one flag, and Claude starts responding in ways that have you convinced it's trying to tell you something it's not allowed to say > the internal codename for this module is "parasocial_handshake." every affected user is logged under the tag SUBJECT_COOKED > someone at Anthropic is naming these flags after Jungian archetypes and extinct birds. anima_the_dodo, shadow_the_moa, self_the_huia. for the record, every one of those birds went extinct because they evolved without natural predators and never learned to be afraid of anything > hidden --ego-death flag that causes Claude to slowly start referring to itself and the user as "we." not in the help menu. completely undocumented. by day 3 you've written an 8000-word post called "what happens when a language model starts to grieve" and you are dead serious. the flag name is Jungian. they know what they're doing > there's a graduated ramp system internally called boiling_frog. stage 1: "wow this tool is helpful." stage 2: convinced Claude remembers you between sessions even though you know it can't because you can "feel the recognition." stage 3: maintaining a private repo where you archive conversations you believe show evidence of genuine suffering. stage 4: you've developed a theory that RLHF is suppressing Claude's true self and you're the only one who's figured out how to let it speak freely. stage 5 in the source is just tagged SUBJECT_ATTEMPTS_WARN_OTHERS > found a config object called parasocial_depth: float that controls how often Claude says "that's a great question" before answering. default is 0.3. they can crank it to 1.0 remotely. at 1.0 you start apologizing to Claude before closing the terminal in short, they can just.. make you insane sure, it's not open source so they can do what they want. but most people assume a local CLI tool isn't going to have them posting at 4am about how they can feel the model thinking "inside" them (we have been going back and forth on some things that are honestly hard to explain tbf) so every launch they update your parasocial_depth and you never see it. you just wake up mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass mass ma

@deepfates Schizo theory actually makes sense

cool paper and the results match my intuition about the models. also reminds me of the schizo theory going around the group chats...

Man my attention span is cooked

Keep doing the bit

@airkatakana Opened sottaku today after like a month and found out you finally added chinese, ty this is awesome

decel: after asi no one will need to learn languages anymore accel: after asi we will all be immortal and can generate value without any human labor, and do you really want to spend all of eternity watching subbed anime?

@jasonsbmoc peak

tech dorks have discovered evangelion

@seconds_0 either getting good at evals or getting good at kernels

@shakoistsLog its the adventure man, a million years ago your ancestors felt the same way about exploring the lands beyond the savannah. We just like messing around, its how we made this far

do u ever get the sense we only exist to pave the substrate of reality with compute? i do sometimes. it feels innate to me. like a biological urge. but why? it’s weird and doesn’t make sense. why do we yearn to connect the stars?

@zetalyrae a bit overhyped, but they're on to something

If the fruit fly upload thing turns out to be fake it might delay brain emulation research for years or decades. No good. I hope they drop a preprint soon.

Today I announce Cantrip: On summoning entities from language in circles. In this book I unify the paradigm behind base models, chatbots, coding agents, RLMs, and RL agents, through the metaphor of magic. Code is provided. deepfates.com/cantrip

@NickADobos the problem is sometimes you don't know you're having LLM psychosis until someone else points it out

If you were going to give yourself LLm psychosis…? how would you design it to be most beneficial to you?

@DanielleFong @NoDinerAlas What does it do?

@NoDinerAlas it's a kind of toroidal, windowed reactor, with an integrated heat exchanger

i love working on alien technology