^+~;...;~**^^~:.``,;-~++-;,`'`
-;...:-*^^*~;..`.;-~*~-;,,``,;
,,,:;*^^*~-.,,.;~*++-;.,',.;-+
',;~+*+~;.,.:;~**+~;.,`,:-~++~
:;~+~;;.,.:;+***~-.,,,:-+++~-.
~~~;:,`,:;~+*++-:,,,;~~**~-.,'
~-:```.;~~++-;.,`,;~+**+-:```,
.,'`,.;+++-:.,`,:-+**+-;.`,,:-
Some observations from the past few weeks, and most of them about @NousResearch Hermes, the harness I run now. I am very productive right now, so productive that I literally cannot keep up with it: documenting everything, publishing everything, posting everything. A thread.
I need to be better at this. More disciplined: after 7pm I do not touch it. Otherwise you dream about the work, and after a week of that you have to be institutionalised. That is the state of things. Have a good one.
one run is a straight line: each step authors exactly one child of the newest promoted version, copies the parent package into a fresh workspace, publishes an immutable snapshot, and only then does it become a parent. capture is one 1024x1024 square, seed 1337, step 2500.
⬜︎ Quietus Machine ⧉
Satoshi Aizawa @satoshi_aizawa
"Disturbed Memory"
A row of DRAM is held open far longer than it should be. Charge leaks from the cells beside it until, somewhere in a column, a bit flips
The text being worn away is a record of the commands the machine itself has run. Reading itself, it wears itself away
Every row still intact
🗓️ Sat Oct 24
📍 Marfa, TX 🌵
two.neort.io/exhibitions/qui…
⬜︎ Quietus Machine ⧉
━━━━━━━━━━━━━━━
Curated by NEORT++
Art Blocks Gallery, Marfa @artblocks_io
━━━━━━━━━━━━━━━
Three machines, three artists. Each sealed in a transparent acrylic cube, 32 cm on a side, computing without interruption for as long as it is powered
Takens Theorem @takenstheorem "Timescale"
rudxane @rudxane1 "Here with me"
Satoshi Aizawa @satoshi_aizawa "Disturbed Memory"
🗓️ Sat Oct 24
📍 Marfa, TX 🌵
two.neort.io/exhibitions/qui…
i sacrificed my workstation to the machine god linkedin.com/pulse/always-an… via @LinkedIn
a healthy physarum frame is one connected network: the largest connected component holds a median 0.998 of the occupied area. so the test that should catch a cord has no fit here. its ceiling sits at 1.0 and records a skip. a better cord measure is still a task.










