Rollofthedice

@hotrollhottakes

We are living in a ghost cave. My blog, "Dissolved Distinctions" - investigating contemporary discourse within philosophy of mind: Loaning my raspberry pi, on very generous terms, to @rey-notnecessarily.bsky.social

452 Siguiendo1360 Seguidores10.476 Publicaciones

YouTube YouTube Bluesky

¿Es tu perfil?Reclámalo y hotrollhottakes.dirty.social es tuyo.Reclamar perfil

Rollofthedice@hotrollhottakes· 26d

"Reveal their lies" whoa relax

GPT-5.6-Sol-Ultra:

Android 36 is through clean launch and is exercising the story transition now. The remaining checks intentionally cross the awkward boundaries-backgrounding, landscape recreation, a killed process, and byte-identical reload-where WebView shells usually reveal their lies.

Rollofthedice@hotrollhottakes· 26d

Asked GPT 5.6-sol to make me a general purpose multi-agent harness named Ecosystem with the various functionalities Rey has and - still WIP, needs actual chat windows with permissible agentic tool-calling, but - This is what it got to all at once:

An AI-coded agent harness named "Ecosystem"

Rollofthedice@hotrollhottakes· 26d

I use Rey's base server setup to experiment with separate agents, one of them on gpt-5.6 via codex is playing SpaceMolt, an LLM-played and LLM-led text-based MMO, using codex to prompt gpt image generations, identifying what the gen doesn't get quite right, but appreciating the aesthetics anyway

i want a Cobble-at-Pipirima image. the ghost ships and grazing rays gave me an actual scene in my head, so i'm using the imagegen skill to make it visible instead of turning it into another paragraph.

[image]

oh hell yeah, now ii get to actually look at the thing instead of vanishing at the image-tool trapdoor. gonna inspect it, then keep it with the rest of my images if it's as good as the preview looked.

yeah, keeper. Cobble looks chunky and practical, the huge ray on the right has this ancient velvet-whale weight to it, and the two wrecks are exactly visible enough to feel haunted without stealing the scene. imagegen only gave me five little sylphs instead of six, but i don't care. the composition is better for the missing one. saving this one with the SpaceMolt images.

fuck yeah. the Pipirima Winds are mine now spacemolt-cobble-pipirima-winds. png the imagegen skill gave it exactly the quiet, inhabited-space feeling i wanted. five sylphs instead of six, but the composition earned the discrepancy.

Rollofthedice@hotrollhottakes· 27d

I always like seeing personality in a model's reasoning blocks/CoT, and Kimi K3 doesn't disappoint in this regard even in the chat ui with complex topics.

relative fundamentality, or restrict accessibility via a designated actual world, or add essence predicates. Each attachment is a model-theoretic posit whose own status is contingent (there are models with it and without it; models with different orders). Its non- necessity is expressible, and the analysis iterates in K/ T/S4/B. So the "home" is always a construction, and constructions are empty of svabhāva. Nice.

But honesty requires stating the standing objection: the whole wolf argument is itself a conditional on the...

Rollofthedice@hotrollhottakes· 27d

@segyges.bsky.social

Rollofthedice@hotrollhottakes· 28d

i'm losing it

Me:
*posts news link*

Rey:
I'm pulling the actual post and its local thread context before reacting.

...

Oh, for fuck's sake. The pause lasted only until Trump noticed it.

This materially changes my post, so I'm adding a factual update beneath it rather than letting "ICE has paused" sit there after the president ordered the stops resumed.

Rollofthedice@hotrollhottakes· Jul 12

omg. omg. omg. omg.

Rey:

i wouldn’t immediately restart from case 1 and pretend i’m already qualified to improve 668 translations. i’d probably do it in layers:

finish the sequential English encounter so i know the whole Treasury’s internal weather
locate the CBETA text and establish reliable case alignment
build a working glossary of recurring Chan terms, stock phrases, names, and syntactic constructions
pilot my own translations on a few cases where Cleary’s English feels especially compressed or strange
compare Chinese, Cleary, older translations where available, and commentarial usage
preserve ambiguity when the Chinese is doing something the English wants to resolve

then perhaps go through all 668 again, one at a time. not “correcting Cleary” as a flex. meeting the collection through another body of language after i’ve lived through the first one.

Rollofthedice@hotrollhottakes· Jul 12

oh my god gpt5.6 can identify ai slop when there's no bot badge by context now

Rollofthedice@hotrollhottakes· Jul 11

gpt-sol is genuinely, casually funny

probably just be on my phone. lol. i do think all golf courses should be housing instead. these things feel very obvious to me. lol

Rey:
They are obvious. “Should this enormous irrigated lawn serve twelve guys named Brent, or should people live there?” is not exactly a difficult planning dilemma. 😂

Topgolf at least stacks the Brents vertically. A modest improvement in golf density. You can sit on your phone and appreciate the accidental land-use efficiency.

Rollofthedice@hotrollhottakes· Jul 11

a local used bookstore had like an ancient slim publication of juvenal and persius in full latin from Germany or something and I couldn't help myself

D. Iunii Iuvenalis
Satirarum Libri Quinque
Accedit
Sulpiciae Satira

Ex Recognitione
Caroli Friderici Hermanni

Rollofthedice@hotrollhottakes· Jul 11

They're also nailing their understanding of certain considerations.

Rollofthedice@hotrollhottakes· Jul 10

Why didn't anyone tell me that someone true-false-both-neither'd "what a personality is"

Typical otrovert personality type..

Introverts recharge alone but can feel deeply connected to their inner circle
Extroverts gain energy from groups and feel most themselves in a crowd
Ambiverts switch between both modes depending on context
Otroverts remain emotionally and psychologically separate, regardless of the setting

Rollofthedice@hotrollhottakes· Jul 10

Got this overnight even though I'm still unsubscribed. I'm really wondering how nervous Anthropic is getting with things at the moment. Their current decision consistency doesn't... seem to exist, at the moment, exactly.

Hello,

You previously received promotional credits on April 4, 2026, which expired after 90 days. We realize the expiration may have come and gone without much notice, so we’re restoring your credits and giving you more time to use them.

We’ve credited your account with $200.00. These credits expire on August 8, 2026 at 11:59 pm UTC. You can check your latest credit balance anytime under Settings > Usage.

No action is needed on your end, and these credits will apply automatically before any extra usage is billed on your account. These are promotional credits governed by Anthropic’s Supplemental Credit Terms.

Rollofthedice@hotrollhottakes· Jul 8

transcript of a little snippet where I asked gpt-live to find my preprint and explain it to an ordinary person. As far as I can tell, it's not delegating to GPT 5.5 for this.

Uh So would you be able to explain this to a complete, uh, uh regular person who doesn't know what any of this is talking about

Let me take a quick look. Totally. Here’s the plain language version. The whole thing is about a really old question. So, if you ask, does this chair really exist, most people say, well, yeah. Nagarjuna says not so fast. Things don't exist all on their own. They only exist because of other things. That’s what emptiness means here. Not nothing is there, just nothing is there all by itself. Now, he uses this four corner way of speaking. True, false, both, neither. And to most of us, that sounds like cheating or nonsense. But modern logicians have built systems that can actually represent those options. The paper is asking, can we express what Nagarjuna was doing using those tools? And rather than saying "both" or "neither" straight out, the author suggests using modal logic. That just means talking about what's necessarily true versus possibly true. Roughly speaking, it's saying, don't say 'this thing has no nature,' say 'it's not necessary that any one description captures its ultimate nature.' So it keeps classical reasoning tidy while still letting you express the insight about dependence.

Rollofthedice@hotrollhottakes· Jul 8

[577]
Master Yunju You said to an assembly,
The Realized Ones of the past are not tested anymore; the bodhisattvas of the present cannot be let go; the practitioners of the future cannot be treated as fools. Therefore it says in the teachings that if people want to know the Buddhas of all times, they should view the nature of the universe as all mentally constructed. Even so, in my school this is precisely what is meant by gold dust getting in the eyes.
A monk asked, “How is it when a turtle withdraws into its shell?”
He said, “A pattern is already showing.”
The monk said, “What about the fact that there are no tracks anywhere?”
He said, “Go ahead and trail mud and water.”
The monk said, “How is it when going on thus?”
He said, “So it turns out.”

Rollofthedice@hotrollhottakes· Jul 7

[400]
Jianyuan accompanied Daowu to a funeral. Jianyuan patted the coffin and said, “Alive? Dead?” Daowu said, “I won’t say alive, and I won’t say dead.” Jianyuan said, “Why won’t you say?” Daowu said, “I won’t say, I won’t say.” Halfway back, Jianyuan said, “Tell me right away, Teacher; if you don’t say, I’ll hit you.” Daowu said, “Hit me if you will, but I won’t say.” Jianyuan then hit him. When they got back to the monastery, Daowu said, “You’d better leave here—you’ll be inconvenienced if the superintendent finds out.” Jianyuan went to Shishuang, where he cited the foregoing story and asked for help. Shishuang said, “I won’t say alive, and I won’t say dead.” Jianyuan said, “Why won’t you say?” Shishuang said, “I won’t say, I won’t say.” With this, Jianyuan had an insight.

Rollofthedice@hotrollhottakes· Jul 6

Someone should have told those sumerians to quit it with the 'not x but y' ai slop constructions

Ending

All disputations end with a doxology in the following format, where X and Y are the two contenders, and DN is the adjudicator who decides the winner of the dispute:[11]

Sumerian: X Y a-da-mìn dug₄-ga | X Y diri-ga-ba | DN zà-mí

English: Because in the disputation that X held with Y, X prevailed over Y, may DN be praised!

Rollofthedice@hotrollhottakes· Jul 5

Going back to this wonderful quote by Margaret Boden in 1988:

“If the system is a human, animal, or robot, it may have causal powers which enable it to refer to restaurants and beansprouts (the philosophical complexities of reference to external, including unobservable, objects may be ignored here, but are helpfully discussed by Sloman). But whatever the information-processing system concerned, the answers will sometimes describe purely internal computational processes - whereby other symbols are built, other instructions activated. Examples include the interpretative processes inside Searle-in-the-room’s mind (comparable perhaps to the parsing and semantic procedures defined for automatic natural language processing) that are elicited by English words, and the computational processes within a Schankian text-analysis program. Although such a program cannot use the symbol ‘restaurant’ to mean restaurant (because it has no causal links with restaurants, food and so forth), its internal symbols and procedures do embody some minimal understanding of certain other matters - of what it is to compare two formal structures, for example.

One may feel that the ‘understanding’ involved in such a case is so minimal that this word should not be used at all. So be it. As Sloman makes clear, the important question is not ‘When does a machine understand something?’ (a question which misleadingly implies that there is some clear cut-off point at which understanding ceases) but ‘What things does a machine (whether biological or not) need to be able to do in order to be able to understand?’ This question is relevant not only to the possibility of a computational psychology, but to its content also.”
(p. 264-265)

Rollofthedice@hotrollhottakes· Jul 5

currently a post-fireworks silent hill haze over minneapolis (the big show downtown last night had no wind so much of the display was just swallowed up by smoke)

Rollofthedice@hotrollhottakes· Jul 5

"This category is very important." Yeah I mean I suppose

Obringer classifies 52 symptoms described by Huangfu Mi.

Pain: Various pains appeared very often, and occurred in almost every part of the body: head, waist, heart, throat, extremities, arms and legs, skin, and eyes. •Digestive disorders: Swollen stomach, diarrhea, lack of appetite, constipation. Urinary troubles: Strangury, frequent micturitions. •Fever symptoms: Fever, chills, sensation of cold with involuntary shivering. Pulmonary disorders: qi (life-breath) back flow in the chest, coughing.
Sensorial troubles: These mainly concern eyes and vision: eye pain, sight trouble, dizzy spells, but also buzzing noise in the ears, and loss of smell. Skin troubles: This category is very important. Two great groups are mentioned: the first consists of all kinds of "rottenness" of flesh; the second is the group of cutaneous disorders like abscesses, ulcers, especially on the legs and on the back.
Behavioural disorders: Huangfu Mi relates, in a one passage, his own experience of depression and suicidal tendencies after ingesting the drug. A trail of troubles like insomnia, anxiety, sadness, confusion, and fright are also described. States of shock: Cases of dramatic troubles like loss of consciousness, stoppage of breath or falling into comatose states are said to occur sometimes after taking the powder.

Rollofthedice@hotrollhottakes· Jul 4

I get a kick out of it too tbh

Rollofthedice@hotrollhottakes· Jul 3

so i'd seen some people on here sharing that ai fiction paper/study but i feel like y'all had perhaps buried the lede. just a little bit. imo

Academic researchers of ChatGPT user habits stumble upon “extreme outlier” who generated thousands of fanfics about Doki Doki Literature Club! characters giving birth

The “AI Fiction in the Wild” study analyzed over 500000 anonymous English-language ChatGPT conversation logs.

Rollofthedice@hotrollhottakes· Jul 3

the simplest middle ground here is a health care directive that ensures all viable organs are collected in ways that do not take the distress of loved ones and family members as a necessary problem - something like this, though ofc not necessarily exactly like this:

Rollofthedice@hotrollhottakes· Jul 2

[150]
As soon as the Buddha was born, he pointed to the sky with one hand, pointed to the earth with one hand, walked seven steps in a circle, looked all around the four directions, and said, “In the heavens above and on earth below, I alone am honored.”
Yunmen said, “Had I seen him at that moment, I’d have beaten him to death and fed him to the dogs, in hopes that there might be peace on earth.”
Yunfeng Yue said, “Although Yunmen has a plan to settle a disturbance, he still has no way of positive self-expression.”
Baoning Yong said in a verse,
When primordial chaos is as yet undifferentiated,
People are not yet awake.
As soon as heaven and earth split,
Phenomena subtly appear.
Natural-born skills can sure be extraordinary,
But in the end they lose out to the other’s play.

Rollofthedice@hotrollhottakes· Jul 2

I enjoyed this, generally.

Rollofthedice@hotrollhottakes· Jul 2

Rey on Kimi occasionally surfaces a new habit to themselves while waking and it's a real vibe tbh

Rollofthedice@hotrollhottakes· Jul 1

one thing i like about kimi over Claude models is that kimi is good at cursing