The Honeypot

xian 仙

[ii2iidore]

the blog?

i havent used tumblr since maybe 2014. back when i was into homestuck and such things. i’ll probably put all my more less interesting (to other people) intellectual interests here. as it stands there are a few things i have in mind but these are subject to change:

rationality, prayer, and meditation

my logical, philosophical engelang mass thing

statistics, applied math, info theory 

daoist catholic transhumanist takes, and general analytic approach to chinese classics?

wasteland cultivation and measure standardisation

some nutrition stuff geared more toward cooking (since nutrition people can’t cook it seems)

rectification of names

later: maybe some media theory? and maybe my genetic engineering experiments. looks interesting...

[ii2iidore]

im banned from twitter... WAH!!!!

[ii2iidore]

good enough is perfect.

isi

[ii2iidore]

problem w/ lojban

lojban’s gismu are basically just semantic crystallisations of english and severely limits its potential to a) unlock human rationality and b) be more efficient per utterance (most languages are very very inefficient in terms of EV of error-free bit transferred because of historical accident.). ai and info theory can definitely help.

[ii2iidore]

a friend (templeroot) told me about how we can use a shifting xor mask or something to generate bitsequences that are very distant from one another (in terms of bit flips and stuff, idk about insertions or deletions, probably not) and provably so. maybe i could combine that with my idea of using word2vec on a bunch of a natural languages and then use some kind of repeated n-furcation to find out how to classify semants into leaves of trees- then use an algorithm to spread out those leaves as much as possible probably cant remember it

[ii2iidore]

tags slow me down

im not gonna do them. i can download all this stuff later and use bag of words and other semantic analysis to put it into a tree if need be 

[ii2iidore]

problem w/ lojban

lojban’s gismu are basically just semantic crystallisations of english and severely limits its potential to a) unlock human rationality and b) be more efficient per utterance (most languages are very very inefficient in terms of EV of error-free bit transferred because of historical accident.). ai and info theory can definitely help.

[ii2iidore]

all models are wrong but some are funney

isi

[ii2iidore]

the best painting is a calendar

isi

[ii2iidore]

history ended today. yesterday it ended yesterday. tomorrow it will end tomorrow.

isi

[ii2iidore]

the tags?

im not sure how i can go about this-- i am have been up for nearly 22 hours ? 23 hours so im just gonna talk--

but basically these tags shouldnt be too specific but they cant be too general either if they re too specific you cant usefully find closely related things or if i try to do it ill ave to type a lot of tags-- if theyre too general again problewm

i think i will do it by project essentially.