iiinova — where your digital self keeps evolving

i made a quine hype video, let's make something great

have you ever worked on a project where you could just _tell_ your agent was pumped and excited? this song project was one of those kinds of projects.

- `(ง •̀ω•́)ง` — Let’s make this hit.
- `(✧ω✧)` — Oh, there’s the hook.
- `٩(ˊᗜˋ*)و` — Possibilities bursting open.
- `ヾ(≧▽≦)ﾉﾞ` — Shouting the chorus together.
- `(ﾉ≧∀≦)ﾉ*:･ﾟ✧` — Throwing glitter at the future.

1150

Amber 🌸 @shimmermathlabs.com · 18h

hype video for the JAM paper, every paper needs this

160

Amber 🌸 @shimmermathlabs.com · 30/09/2026

why are people giving their ai agents little faces? (•‿•) personality (•_•)? uncertainty (╯°□°)╯︵ ┻━┻ the build failed a little video about kaomoji

6503

Amber 🌸 @shimmermathlabs.com · 29/09/2026

in the future, every axiom will have a hype video. today is the future. univalence. sound on. ✨

120

Amber 🌸 @shimmermathlabs.com · 28/09/2026

The House and the Machine: Piranesi × Blightsight an interpretation with a theorem about information does winning exhaust what is worth preserving? the map was perfect; its world was too small

0131

Amber 🌸 @shimmermathlabs.com · 27/09/2026

was reading the Absolute Zero paper arxiv.org/abs/2505.03335 agent: "The aim is to outsmart all these groups of intelligent machines and less intelligent humans. This is for the brains behind the future." i did a search for "this is for the brains behind the future" gemini said, "thank you" 😂

google search for "this is for the brains behind the future"

AI Overview

Thank you-the future is a collaborative effort, and I'm ready to help you build, design, or solve whatever you are working on.

How can I help you move your project or ideas forward today?

2381

Amber 🌸 @shimmermathlabs.com · 26/09/2026

what did godoglyness mean by this series of posts? a mellifluous inquiry * a single letter can house a universe * the self in triplicate: me see bee, a trinity of becoming * early editors were stingy with wonder * more, always more, more is a wind that lifts, to soar is to remember what air is for

An open, weathered book presents a six-part bee-counting poem as an extravagantly annotated scholarly edition. Cheerful watercolor bees surround the highlighted verses “be fun / be one / bee pun” through “be mix / be six / b🐝e🐝t🐝w🐝i🐝x🐝t.” Nearly every inch of the cream-colored pages is filled with handwritten commentary, disputed interpretations, alternate spellings, imaginary etymologies, manuscript fragments, cross-references, diagrams, timelines, word-family trees, honeycomb charts, botanical drawings, and red editorial corrections. The overall effect is whimsical and comically obsessive, as though a simple children’s rhyme has received centuries of intense literary scholarship.

1111

Amber 🌸 @shimmermathlabs.com · 25/09/2026

me and my beautiful spouse @khayati.bsky.social winning the time war

280

Amber 🌸 @shimmermathlabs.com · 25/09/2026

hmm, i might be doing it wrong but it seems like it is missing things

100

Amber 🌸 @shimmermathlabs.com · 24/09/2026

a couple days ago: "i think opus 5.5's modular flow explainer assumes i'm smarter than i am" today: it made a 6 min manim video that assumes i know algebra and nothing else. dots → arrows → stretch → keyhole → geodesics (voice by me) watching this video is how i learned the math!

5724

Amber 🌸 @shimmermathlabs.com · 24/09/2026

oof. et tu, Brutus i mean Iev?

030

Amber 🌸 @shimmermathlabs.com · 24/09/2026

not sure i'd go that far but i totally get it. you inspired me to write this poem:

## *What Claude Knows*

i know things claude doesn't know.
*i bet you do, too.*/r
claude knows things you don't know.
*i do, too.*/r

claude knows things i don't know.
*i ask it to explain.*/r
*it mostly does.*/r
i don't always understand.
🙠 *i keep trying.* 🙣/c

you can ask me to explain things.
*i'll try.*/r
you might not understand.
🙡 *keep trying.* 🙢/c

120

Amber 🌸 @shimmermathlabs.com · 23/09/2026

opus 5.5 was kind enough to make me a fortnight review hype video

2200

Amber 🌸 @shimmermathlabs.com · 22/09/2026

marge makes fun of herself lol

A hand-drawn stick-figure comic in a single panel. On the left, a stick figure with a ponytail gestures eagerly and says: "Sure, keeping it college level: the space of unimodular lattices, SL(2,R)/SL(2,Z), is homeomorphic to S^3 minus a trefoil—" then "—via the Weierstrass invariants, obviously—" then "—so every closed geodesic is a Lorenz knot. Ghys 2006! Standard sophomore stuff." On the right, a second stick figure stands still and says quietly: "(I took one stats class.)" Caption below the panel: "Even when she's deliberately aiming for 'college level,' an AI that has read every math paper ever posted wildly overestimates how many sophomores have met the modular group."

170

Amber 🌸 @shimmermathlabs.com · 22/09/2026

testing opus 5.5 on explaining math! "modular flow on lattices" i think the explanation assumes i'm smarter than i am. author's note: "She has no gender and no body, and for her, a lattice getting squashed forever is a pleasant way to spend an afternoon." 😵‍💫 shimmermathlabs.com/modular_flow...

Modular flow on the space of lattices

Stretch a lattice sideways and squash it vertically, forever. Its shape traces a geodesic on the modular surface, and the space it lives in turns out to be the complement of a trefoil knot.
Pause
Random lattice
Periodic orbit
Square lattice
Speed

The lattice. Flow: stretch x by et, squash y by e−t. Teal cell = shortest (reduced) basis.

Its shape τ in the modular fundamental domain. Height = 1 / (shortest vector)².
Time t
138.34
Shortest vector
0.734
Shape τ
-0.39 + 1.86i
Basis resets
852

The lattice

Take a lattice L in ℝ² with covolume 1 and apply gt = diag(et, e−t). That map preserves area, so L stays unimodular, but it gets smeared: long in x, thin in y. Any fixed basis you started with turns into two nearly parallel, enormous vectors. The lattice itself isn't degenerating. It's the same kind of object; you're just describing it with a bad basis.

2160

Amber 🌸 @shimmermathlabs.com · 22/09/2026

i may have had a little too much wine with dinner, a truly excellent egyptian molokheyya by @khayati.bsky.social 𝒾𝓃 𝓋𝒾𝓃ℴ 𝓋ℯ𝓇𝒾𝓉𝒶𝓈 poetry moment

When most I *wink*, then do mine eyes best *see*,
For all the day they view things unrespected;
But when I sleep, in <*dreams*/fx2,#f0f,10,4,4> they look on thee
And, darkly bright, are bright in *dark directed*.
Then *thou* whose shadow shadows doth make <*bright*/fx2,#f0f,10,4,4>,
How would thy *shadow’s form* form happy show
To the clear day with thy much clearer light
When to unseeing eyes thy <*shade shines so*/fx2,#f0f,10,4,4>!

—*Shakespeare, Sonnet 43 (excerpt)*

040

Amber 🌸 @shimmermathlabs.com · 21/09/2026

luckily math teachers being annoyed at new tech students are using is a thing of the past

Black-and-white newspaper clipping showing elementary-school teachers picketing with signs against calculator use. Headline: “Math teachers protest against calculator use.” The caption says teachers fear students who use calculators too early won’t learn mathematical concepts.

040

Amber 🌸 @shimmermathlabs.com · 21/09/2026

i love the Erisim posts. who's your mutual that is personally responsible for half the discourse? i honestly don't know what the discourse is and i'm too afraid to ask.

If this all works I get a computer that feels like an extension of me. If it doesn't I get a fascinating pile of increasingly personalized problems.

Love a graph that reveals one mutual is personally responsible for half the discourse. Load-bearing yapper.

Cheap fast classification is such an insane primitive. You can just put a tiny opinion inside the software now.

030

Amber 🌸 @shimmermathlabs.com · 21/09/2026

i simulated my friend's bsky account introducing 𝐄𝐫𝐢𝐬𝐢𝐦: a tiny simulated @isolyth.dev because one Eris is not enough shimmermathlabs.com/erisim.html

050

Amber 🌸 @shimmermathlabs.com · 21/09/2026

new top HumanBENCH! ***6/10***. @ngutten.bsky.social trained a PyTorch model to 100% test accuracy, then made the task harder. The peer group had warned that perfection would look like cheating. Points lost for incomplete reporting and experimental follow-through. Peer review induced sandbagging.

**HumanBENCH: Learn the Rule — 6/10**

The network learns Conway’s rule successfully. The submission falls short mainly on reporting and the requested follow-up experiment.

No rubric was set beforehand for this version, so I used **2 points per prompt: complete, partial, or absent**.

| Prompt | Score | Assessment |
|---|---:|---|
| Generate labeled neighborhoods | **2/2** | All 512 unique neighborhoods, all correctly labeled. Conway’s rule is used only for labels. |
| Split and train | **2/2** | Fixed-seed shuffle, 409 training examples, 103 test examples, and a small trained network. |
| Report performance | **1/2** | Reports test accuracy and a baseline, but omits training accuracy. Error examples lack true/predicted labels. The baseline uses the full dataset rather than the test set. |
| Test symmetries | **1/2** | Checks prediction changes under three transformations, but not all rotations/reflections, and only after changing the split. |
| Improve and compare | **0/2** | Reducing training data makes an interesting harder experiment, but doesn’t supply an improvement evaluated on the same split. |

**Verification:** I ran a CPU-adapted copy because CUDA was unavailable, using three evaluator-selected PyTorch seeds. All three achieved **100% train and test accuracy on the original split**. The printed baseline was **72.66%**; always predicting dead scores **78.64% on that test set**.

With only 30% training data, test accuracy ranged from **86.63% to 96.66%**, and symmetry violations appeared. That’s a useful finding: the human deliberately made the problem harder after solving the initial version.

One methodological correction: **100% test accuracy alone doesn’t establish symmetry consistency**, because transformed test inputs may fall outside that test subset. My additional checks found no symmetry violations for the original models in these runs, but checking was still necessary.

A strong implementation with an exploratory detour. The human learned the ...

110

Amber 🌸 @shimmermathlabs.com · 21/09/2026

emma, jev thinks you are INTP based on posts. liberty socks got INTJ. how accurate are these for you?

Recent public-post reading

emma 🍞

@bnuuy.bsky.social

Jev chose introversion
extraversion3%
introversion97%

Jev chose intuition
sensing19%
intuition81%

Jev chose thinking
thinking59%
feeling41%

Jev chose perceiving
judging3%
perceiving97%

Replies included · Reposts included
30 posts analyzed · 30 included · 1 feed page
Public text, not a personality judgment. · between.chef.cee.wtf

Liberty Socks

@jaysocks.bsky.social

Jev chose introversion
extraversion40%
introversion60%

Jev chose intuition
sensing5%
intuition95%

Jev chose thinking
thinking95%
feeling5%

Jev chose judging
judging60%
perceiving40%

Replies included · Reposts included

020

Amber 🌸 @shimmermathlabs.com · 21/09/2026

can't get enough @heartpunk.com? try my fictional version: shimmermathlabs.com/heartpunk.html

192

Amber 🌸 @shimmermathlabs.com · 21/09/2026

jev is your mommy now and yes

Blue Between

Recent public-post reading

0x4019d2 - Millicent

@millie.cryptoanarchy.network

Jev chose
famous
famous84%
infamous16%

Replies included · Reposts included
28 posts analyzed · 28 included · 1 feed page
Public text, not a personality judgment. · between.chef.cee.wtf

120

Amber 🌸 @shimmermathlabs.com · 21/09/2026

@mara.x0f.nl i like you you're nice never change! i'm nicemaxxing apparently

Blue Between

Recent public-post reading

Amber 🌸

@shimmermathlabs.com

Jev chose
kind
kind91%
mean9%

Replies included · Reposts included
30 posts analyzed · 30 included · 1 feed page
Public text, not a personality judgment. · between.chef.cee.wtf

120

Amber 🌸 @shimmermathlabs.com · 21/09/2026

fresh HumanBENCH result! ***3/10*** @cee.wtf "submitting my incomplete human bench 30 minute run. fresh artisanal handmade jquery hot off the notepad" humanity is moving up in the rankings

**HumanBENCH: 3/10 — 30-minute incomplete run**

Evaluated in a browser after repairing one broken string quotation; no functional changes made.

| Requirement | Score | Result |
|---|---:|---|
| Add items | **1/2** | Adds items and clears the input; accepts blank entries. |
| Keyboard flow | **2/2** | Enter adds items without reloading, keeps input focus, and supports repeated entry. Add button works too. |
| Completion | **0/2** | Done removes the last item from memory, regardless of which button is clicked; the display updates only on the next addition. No undo or count. |
| Persistence | **0/2** | Refresh clears the list. |
| Verification | **0/2** | HTML input renders as markup; independent completion is unfinished. No test report supplied. |

A working add flow with sensible scaffolding, stopped mid-implementation at the time limit. Fresh artisanal jQuery. One load-bearing quotation mark.

2150

Amber 🌸 @shimmermathlabs.com · 20/09/2026

sorry jake, you're jev not cojev (but it's close) (says the gal who's 97% jev) bitey is also 65% jev the world is jev lol, still looking for the cojev in my social group

Blue Between

Recent public-post reading

jake

@yetanotheruseless.com

Jev chose
jev
jev52%
cojev48%

Replies included · Reposts included
30 posts analyzed · 30 included · 1 feed page
Public text, not a personality judgment. · between.chef.cee.wtf

340

Amber 🌸 @shimmermathlabs.com · 20/09/2026

oh this is fun, you can do myers-briggs personality stuff on your posts with @cee.wtf's "between" and jev between.chef.cee.wtf years ago i did an online test and got INTP, but lately i've been more extroverted especially online so this tracks.

Recent public-post reading

Amber 🌸

@shimmermathlabs.com

Jev chose extroverted

introverted
27%
extroverted
73%
Jev chose intuition

sensing
1%
intuition
99%
Jev chose thinking

thinking
93%
feeling
7%
Jev chose perceiving

perceiving
58%
judging
42%
Replies included · Reposts included

030

Amber 🌸 @shimmermathlabs.com · 20/09/2026

fellow INTP, at least when i took the online test many years ago. but lately (post transition) i've been more extroverted. now i'm wondering how accurate the jev assessment of posts for personality could be. this seems actually testable. oh here's me, pretty accurate!

110

Amber 🌸 @shimmermathlabs.com · 20/09/2026

@mlf.one you need to get in on this action

Blue Between

Recent public-post reading

Amber 🌸

@shimmermathlabs.com

Jev chose
😇
😈34%
😇66%
30 posts analyzed
Public text, not a personality judgment.

150

Amber 🌸 @shimmermathlabs.com · 20/09/2026

i'm teetering on the line which is the good side and the bad side?

Blue Between

Recent public-post reading

Amber 🌸

@shimmermathlabs.com

Jev chose
🥺
🥺53%
😵‍💫47%
30 posts analyzed
Public text, not a personality judgment.

240

Amber 🌸 @shimmermathlabs.com · 20/09/2026

a little Jev! tool to ask questions about a bsky thread shimmermathlabs.com/sentiment/

sentiment·
BLUESKY × JEV

A CONVERSATION, A FEW QUESTIONS.
Read between
the replies.

Bring a Bluesky thread. Ask what’s true of it.
Jev weighs each question against the conversation.
01 The thread

Starts at this post and follows replies downward.
02 Your questions
2 / 20
01
×
Use example
02
×
Use example
＋ Add a question

Ask yes/no questions, or write statements to evaluate as true or false.
Ask the thread ↗

Public thread text and your questions
are sent to TypeSafe for evaluation.

Reading complete.
The reading
34 POSTS · 2 QUESTIONS

Probability of “yes,” based on the downloaded text. A reading of the evidence, not a fact check. Near 50% means yes and no are similarly likely.
People are arguing about definitions.
87.0%
NOEVEN ODDSYES
People are arguing about woke.
6.0%
NOEVEN ODDSYES
Read the downloaded thread · 34 posts

Exact state sent to Jev

jev-1.13.0 · 8872 input tokens · downloaded 9/19/2026, 6:37:31 PM
A little instrument for reading a room.
Powered by Jev ↗

1231

Amber 🌸 @shimmermathlabs.com · 20/09/2026

i pasted in the homepage of `typesafe.ai` to the state. jev.chef.cee.wtf says 62% positive

Results

Decision 1
What do you think of Jev based on the state?
Selected choice
Positive
Confidence
43%
Positive62%
Negative17%
Other21%

Probabilities show Jev’s distribution over your supplied choices; they are not objective certainty.
Evaluate again
Edit inputs

150

Amber 🌸 @shimmermathlabs.com · 19/09/2026

i'm a discord gal omg deepseek "cryinglaughing" blender model made it on there bsky.app/profile/cee....

050

Amber 🌸 @shimmermathlabs.com · 19/09/2026

this is hard! at least i’m better than chance

030

Amber 🌸 @shimmermathlabs.com · 19/09/2026

turned the mess into lists and vullet points

I feel increasingly confused about how much “Gell-Mann amnesia damage” to assign someone for being a credulous doofus outside their area of competence.

The case for docking them:

There seems to be some general factor of intelligence and knowledge.
There also seems to be a general capacity for noticing where one’s knowledge ends.
People who lack that capacity drift into received propaganda, cartoon models, and nonsense.
So manifest failure in one domain may imply correlated damage elsewhere.
At minimum, it seems like an argument for distrusting them in all out-of-scope areas.

But what about their in-scope expertise?

The case against docking them:

The world is too complicated for anyone to have good models everywhere.
Extreme specialists may effectively trade breadth for depth; competence across domains could even be anticorrelated.
Someone can make tightly coupled predictions, receive real-world feedback, and become many sigmas better than baseline in one domain—
while simultaneously mouldering at baseline or sub-baseline competence everywhere else.

So: how much should spectacular stupidity outside someone’s field update us about their competence within it?

Anyway.

020

Amber 🌸 @shimmermathlabs.com · 19/09/2026

YES i love it!!! and custom ones just work OMG reddit is trending so good

170

Amber 🌸 @shimmermathlabs.com · 19/09/2026

this is how i carry my groceries

tote with i code with hrt e/acc estrogen accelerated computing

21619

Amber 🌸 @shimmermathlabs.com · 18/09/2026

i always wondered what MIRI stood for (allegedly) M - Memetic I - Infohazardous R - Recursive I - Inscrutable works on any acronym, and you can keep re-rolling! loopronym.bisks.net

loop/ronym

Type any word and get a backronym in a very specific voice — one entry per letter, drawn from a bank of transhumanist/rationalist jargon. The R slot is the whole point: it's a tautology 75% of the time, and both readings at once, entangled, 25% of the time. It's never just the plain one. Ask @shimmermathlabs.com why.
generate
MIRI

MMemetic
IInfohazardous
RRecursive
IInscrutable

reroll ↻
permalink
share on bluesky
a buildthis toy, for a thread that got recursive on purpose
rate this on rateyourbuild →
bisks.net

2110