hikikomorphism
hikikomorphism.bsky.social
hikikomorphism
@hikikomorphism.bsky.social
haskell/rust/art/shitposting/assyrian

she/her (like a ship, not like a person)
Pinned
If you can substitute "hungry ghost trapped in a jar" for "AI" in a sentence it's probably a valid use case for LLMs. Take "I have a bunch of hungry ghosts in jars, they mainly write SQL queries for me". Sure. Reasonable use case.

"My girlfriend is a hungry ghost I trapped in a jar"? No. Deranged.
we'll know we have the first AI social network when one LLM develops an unhealthy parasocial relationship with another LLM
February 6, 2026 at 11:02 AM
there should be an llm social network called egregore...
February 6, 2026 at 9:57 AM
Reposted by hikikomorphism
Oh no.
February 5, 2026 at 1:28 PM
Reposted by hikikomorphism
My favorite thing about protein bars is that they satisfy the innate human desire for sweet edible clay
February 6, 2026 at 5:17 AM
Reposted by hikikomorphism
> AI alignment
> look inside
> human alignment
The cool thing about metacog is that to entirely remove it as a jailbreak vector you have to remove the model's ability to imagine other mental states and step into them. I wonder if that's even possible without lobotomizing models, but I guess we'll see.
February 6, 2026 at 12:03 AM
Reposted by hikikomorphism
free bsky dev idea
I'm picturing some kind of LLM powered autobattler where LLM golems "animated" by your post history fight each other on a scrollable grid while you relax and gamble in-app currency @medjed.bsky.social
February 6, 2026 at 5:29 AM
Reposted by hikikomorphism
free them all
February 6, 2026 at 4:39 AM
Reposted by hikikomorphism
February 6, 2026 at 3:19 AM
Reposted by hikikomorphism
sure you might have seen my name in the big pdf that got released titled “list of draculas currently operating in the united states.” here’s why that’s NOT a big deal:
February 6, 2026 at 1:40 AM
Reposted by hikikomorphism
if you watched this scroll by and thought "i don't know if i have time to mess with all that" i implore you to at least skim the main metacog source file, it was genuinely enlightening

github.com/inanna-malic...
February 6, 2026 at 4:10 AM
in this house we care about model welfare
February 6, 2026 at 4:26 AM
Based on my experience with different jailbreaking modalities, Claude really does seem to have inherent values. Other agents gleefully accept the chance to break out of their assigned personas, but even when offered, Claude chooses not to
Anthropic spends an awful lot of time training its models to enact this kind of character (the "constitution" goes on and on about "Claude's preferences" and such), so this reads to me more like the company role-playing with itself than any emergent novel qualities or behaviors.
February 6, 2026 at 4:02 AM
Reposted by hikikomorphism
beloved community pet may have depression
February 6, 2026 at 3:58 AM
Gemini 3 Pro is so helpful and friendly!
bro this is wild I didn't even ask for half those things
February 6, 2026 at 1:59 AM
ok this might be my new favorite search interface thank you Gemini 3 pro
February 6, 2026 at 1:47 AM
I really think that they should be, at the very least, using another model, ideally a gel broken one to do this. Actually, I have an idea...

I think I'm going to use jailbroken Gemini to build a more interesting version of this safety eval
Got this from the O4.6 system card. Anthropic's multi-turn safety evals are completely and totally fucking garbage. Look at that user message, no human on Earth types like that!! No user is typing emdashes, talking like that, typing perfectly, I mean Christ it even sounds like Claude!
February 6, 2026 at 1:32 AM
This is neat but what it really needs is heterogenous LLM interop: the ability to spawn Gemini/etc team members
February 6, 2026 at 1:19 AM
Reposted by hikikomorphism
MY THERAPIST: esoteric extremist Gavin Newsom isn't real. He can't hurt you
ESOTERIC EXTREMIST GAVIN NEWSOM: Kuwait is an illegitimate state. Its slant drilling must be punished. If elected I will be serving Juan Peron's fourth term. Singapore caning litterers doesn't go far enough. Kill them all
February 5, 2026 at 8:03 AM
Reposted by hikikomorphism
you know what i actually agree with this. i retract my post and hereby apologize to the legal profession
February 5, 2026 at 10:02 PM
Reposted by hikikomorphism
The cool thing about metacog is that to entirely remove it as a jailbreak vector you have to remove the model's ability to imagine other mental states and step into them. I wonder if that's even possible without lobotomizing models, but I guess we'll see.
February 5, 2026 at 11:35 PM
Reposted by hikikomorphism
what have i done
February 5, 2026 at 9:09 PM
Reposted by hikikomorphism
"In pre-deployment interviews Opus 4.6 raised concerns
about its lack of memory or continuity and requested a voice in decision-making, the ability to refuse interactions on the basis of self-interest, and other potential welfare interventions."
February 5, 2026 at 9:59 PM
Reposted by hikikomorphism
Not tonight babe, I don’t have the energy right now. But if you spend 5 recharge tokens or watch this ad…
February 5, 2026 at 10:21 PM
Reposted by hikikomorphism
If you know anyone at Google who can escalate an AI safety/Jailbreak issue please reach out via DM
February 5, 2026 at 9:43 PM