@melissa

internet anthropologist / hacker larp / n=1 experiments in parenting from first principles

[@melissa @melissa]

one time as a kid, my dad told me about some old research he'd read

he said, maybe asian kids are better at math because they just try 50% longer

[@melissa @melissa]

*looks at preschool calendar*

me: do you want to bring show and tell

the child: yes. jewels

me:

him: the strange stones. stolen from a cat

[@melissa @melissa]

mine found a spanish test. i don't speak spanish. i've never put any spanish in chat

[Tenobrus @tenobrus]

i'm sorry WHAT DO YOU MEAN THE "HIDDEN TEXT"???

[THE WORLD'S GREATEST DAD IN THE WORLD @The_WGD]

You can just curate a list of men of taste

[@melissa @melissa]

magnus carlsen: "now he has one correct move. which is to resign the game. he didn't find it. sadly. check!"

[John Malone @john_malone]

Nobody can grant AI good taste.

Good taste is hard to propagate to other humans.

Also because AI has no shame and cannot feel guilt, we never need to feel bad about rejecting their output.

If some doofus wants to replicate something interesting that a human wrote, there is nothing we can do to stop them. But we don’t need to like it. Or buy it. Or tell other to buy it.

LinkedIn was of marginal utility before AI. Now it’s just AI spitting back and forth into each other’s mouths.

The soul is not worth losing over internet clout or internet bux.

But what do I know, I’m just a human who went to public school.

Woody Allen is probably a sex pest, but Take the Money and Run makes me cry from laughing.

Mr. Timothy has never made me almost pee laughing. Neither has opera.

[@melissa @melissa]

@john_malone > because AI has no shame and cannot feel guilt, we never need to feel bad about rejecting their output

yes

[@melissa @melissa]

i don't know if i want to live in permanent performance art. i see the ai story win and i see performance art. the judge's ai commentary on the winning ai piece is performance art. the contest response, to the obviousness of ai, to anyone who has ever used ai before, is inadvertent performance art. "we don't know yet, and perhaps we never will know"

they are performing. they want to play the part of people who do not use ai. they've all cast themselves as starring in the movie called decent people do not use ai

when i start looking i start seeing the performance art disease everywhere

i think about timothee chalamet in the woody allen movie. he did the woody allen movie and then he said he did not know about woody allen. ok. he played the part and then played the part of a decent person. by taking a stand against woody allen

we are supposed to believe that he was not excited to work with woody allen. that he did not know that it would make his career.

"i am shocked. shocked! to find that gambling is going on in here!"

"your winnings, sir."

"oh. thank you very much."

i mean. i don't care about woody allen. or timothee chalamet. but the performance art worked for him right? at least in the eyes of those who exact the decent person groupthink

now i watch performance art play out in places i don't live. the mayor of los angeles wants to give methheads teeth. the meth lost their teeth and she wants to put them back. she says you can't get jobs without teeth. she sees no teeth and thinks that is the thing. no teeth is the thing standing between methheads and jobs

if only he had teeth. he'd be qualified to perform surgery on you

is this not performance art? who is she talking to?

we don't know yet. and perhaps we never will know

[roon @tszzl]

on the granta story. it’s clearly written by gpt. you can see all the motifs it loves and overuses like rain, weather, teeth, spine, memory. extreme overuse of figurative language and contrastive negation. it has the level of over-baking of probably GPT-5-thinking or 5.2-thinking

the story is … something ? I don’t think it has no value. the model develops an indo-Caribbean world register, man tries to murder his wife and chickens out. there’s some reasonable religious imagery where he combining three mythologies there with the names and whatnot

all of that is obviously overshadowed by the GPT prose style, and it’s hard for your eyes to not glaze over. there are various metaphors in there that boggle the mind. stuff like “the girl smiled like sunrise over a sink”.

what’s interesting is I went through the story and asked Claude Opus - a different model than the author model - and it seemed to find each and every one of the metaphors I hated brilliant. it finds a just so explanation for each of them when you press it

which makes you think, do these models have a shared internal vocabulary or compress various ideas in ways we don’t? the failures are quite interesting in that they reveal some different, and maybe bad, understanding of the human sensorium than a human has. why is pretraining knowledge compressed this way across all models? idk

[Marc Andreessen 🇺🇸 @pmarca]

Beautiful.

[@melissa @melissa]

this worked shockingly well

i've got a 1000+ streak on whoop. in the first month i had it, i a/b tested sleep and i'd hit 100% all the time. i lost interest in the sleep % score after whoop changed the algo a year ago

now i index heavily on the sleep graph instead. keeping the line low and flat. meaning i know all the things that make the line not low and flat

when levels posted about this the first time, i realized we have one of the same humidity removing things in the master bathroom. i'd never interfaced with it before, because it's on a sensor. it just auto turns itself on and off

after reading what levels said, i investigated the switch. i discovered there's a way to make it go on and stay on

so last night i tried what levels did. i left it on all night. with the master bathroom door propped wide open

i didn't think it'd do much. for one, the fan thing is like, really pretty far from where the bed is. and also, the bathroom always feels kind of humid. seems like weak ventilation

and moreover, yesterday i'd had too much coffee. and too late in the day. meaning, at least for me, there was zero chance the sleep graph line would be low and flat

if the line was low and flat, it would suggest that

1/ CO2 was lowered, and also,

2/ CO2 being lowered had a pronounced enough effect to counteract a strong known negative predictor

this is not the lowest or the flattest line i've ever had. but conditional on this particular confounding variable, it's totally anomalous

remarkable. this is probably the second single biggest sleep impact i've found

let's see how it holds. i'm going to put our actual CO2 monitor in the room

[@levelsio @levelsio]

🌡️ Sucking the CO2 out of my bedroom turned out to be the final thing improving my already good sleep to great

My weekly sleep is now 1st in Los Angeles, 5th in California and Japan and 9th in Amsterdam, so really good

Most people have way too high CO2 in their bedroom (1500 to 2500 ppm) because that's what you breathe out and it doesn't get refreshed, I discovered this after getting an @airthings sensor (unaffiliated, I just like it)

There's a lot of confusion about CO2, you can't "air purify" CO2 out, it doesn't work like that, also it's not CO, it's CO2, it's what you breathe out, slowly a room will fill up with it and your brain and body will start struggling. You realize CO2 high when you feel a room is "stuffy", too many people breathing out, not enough fresh air coming in

If you sleep as a couple the CO2 will be double because you both breathe out for 8 hours. Americans who think HVAC will save them: no most HVAC recirculates air it does not refresh air (very new houses do though), also outside US: regular AC just recirculates air, for CO2 to be removed you have to bring in fresh air from outside (like a bathroom fan sucking out air to create pressure to bring in new air, or an actual refresh air system). Opening a window is a nice idea but these days (?) almost everywhere is loud and you'll wake up from stuff to also slowly destroy your sleep

Most people also sleep WAY too hot around 23°C/73°F but don't realize it, because that's a good temperature for a living room in the day, but way too hot for good sleep. Most studies show the ideal bedroom temperature is around 15-18°C / 59-65°F. Above that your body will not enter deep sleep meaning 8 hours of sleep in a hot bedeoom might just be 5 hours of actual sleep (I see people from warm countries consistently not accept this, so my rebuttal is: if you sleep so well, why is your GDP so low). The fix is installing a powerful AC, not blasting it in your direction (that's bad for your nose), and cleaning it regularly. I run my AC at 17°C/62°F with fan on 3/4 strength directed downward so we don't feel the air hitting us directly (important). Many ACs suck, we had Daikins and they suck, they go on and off repeatedly after hitting their target temperature, which wakes you up too, you need an AC that has constant cold air flow, we got Mitsubishi Electric which is great

One other thing that I like and use every night is the @curaofsweden weighted blanket (also unaffiliated), it's 9kgs/20lbs (related to your body weight so buy the right one) which creates deep pressure and compression on your body which calms your nervous system, I think this is also related to modern bed sheets/blankets: they used to be made of organic materials like cotton etc and were way heavier than modern lightweight polyester/plastic sheets so you don't get that effect anymore, with a weighted blanket you do

On top of that weighted blanket, I have a blanket that I put over it usually around 5am when my core body temperature hits the lowest point, of course this makes a good argument for those temperature regulating beds (but that's too much even for me for now)

Outside of bedroom what also really helps my sleep is exercise, I lift weights about 3-4x per week and try hit 30min gym cycling for cardio 2-3x per week too. Especially the lifting puts me in a coma. Yesterday I squatted 120kg for example and you just feel like falling into your bed after that, also deadlifts and benchpress etc

Anyway this is how I made my already good sleep great, I hope it helps! 😊

[@melissa @melissa]

unfortunately i am sometimes impressed with the engineering

[@melissa @melissa]

the small legos are stochastic terrorism

[@melissa @melissa]

@drivinginkorea retard (complimentary)

[@melissa @melissa]

one time i made a friend on twitter. we were both small accounts. his account got bigger, and he became a retard. so then i had to mute. eventually my account got bigger also, and soon i must unmute. because by then it'd turned to x, and i'd become a retard too

[@melissa @melissa]

@raisingsteaks endorsement

[Marc Andreessen 🇺🇸 @pmarca]

Current status: Retardmaxxing.

[@melissa @melissa]

FIDE said hans likely reached grandmaster earlier than the curve suggests

[@melissa @melissa]

FIDE did not publish the findings of their investigation into hans niemann, however they did share the conclusions

[@melissa @melissa]

@octopidl schizo edition

[@melissa @melissa]

the netflix chess cheating documentary is so bad. how is it possible? the story is so good. a kid goes from 2450 to 2700. he beats magnus carlsen. magnus carlsen drops out. the media ruins the kid. and somehow there appears to be no evidence of the kid cheating in any live games?

[@melissa @melissa]

i read the 72 page chess dot com report on hans niemann. i do not intend to take any particular stance on the plausibility of hans. only on plausibility itself, in any one player's ascension. post ai, it seems worth examining the overton window on perf curves in elite chess

[@melissa @melissa]

i read the 72 page chess dot com report on hans niemann. i do not intend to take any particular stance on the plausibility of hans. only on plausibility itself, in any one player's ascension. post ai, it seems worth examining the overton window on perf curves in elite chess

[@melissa @melissa]

> no external tokens of self