Havoc Crow

605 posts

Havoc Crow

Havoc Crow

@Havoc_Crow

here Katılım Ekim 2023
33 Takip Edilen6 Takipçiler
Havoc Crow
Havoc Crow@Havoc_Crow·
As I am packing for my move from Japan, I've discovered that wrapping things in clingwrap is insanely fun. I'd wrap the world if I could.
English
0
0
0
2
Havoc Crow
Havoc Crow@Havoc_Crow·
@Saturn_Shuttle Pickup line: "Hey baby, I really need your help in writing my legal brief"
English
0
0
5
828
senseability
senseability@stimpdimpy·
@selentelechia 5yo obsessed with time. gave him an alarm clock for his room but took it back cause he was staying up for hours watching it. a month later realize he’s been sneaking out of his room multiple times a night to go down two flights of stairs and check the time on the basement stove
English
1
0
55
1.2K
🌾🍁🍂 bosco 🍂🍁🌾
for months, my 5yo has been asking "how many minutes until x" and I would tell her "y minutes" and then she'd fall to the floor shouting "Y MINUTES?! THAT'S TOO LOOOOOOONG I HAVE TO COUNT TO SIXTY SO MANY TIMES! 😭😭😭" for any and all values of Y but now she has a wristwatch that also has minutes on it, to learn to tell time so our new routine is constant updates about how close the minute and hour hands are to the relevant numbers on her watch, with commentary like "look! all that time we're waiting is really paying off! Mommy you never need to wonder what time it is again because I have a watch!" I love it
English
19
5
1.6K
46.2K
Havoc Crow
Havoc Crow@Havoc_Crow·
@TraceyRyniec @ObinnaOkpolu Sorry, but this just isn't true. There's been more than enough research on the matter; An LLM is more sophisticated than a Markov chain generator, it has enough inner 'mechanism' to grasp meaning.
English
1
0
0
72
Tracey Ryniec
Tracey Ryniec@TraceyRyniec·
@ObinnaOkpolu That's his point. AI doesn't KNOW it doesn't mean the same thing. It is just stringing together the phrases it finds elsewhere. Story might be a completely different genre or topic. It doesn't know or care.
English
1
0
1
187
Alex Imas
Alex Imas@alexolegimas·
This from @TuhinChakr is brilliant. That prize winning story from Granta? Turns out it's just a bunch of random whole phrases taken directly from existing text on the internet. Tool allows you to trace those n-grams directly to their source, which is mostly random fanfiction. tuhinchakrabarty.substack.com/p/ai-slop-gran…
Alex Imas tweet media
English
37
110
762
298.8K
Havoc Crow
Havoc Crow@Havoc_Crow·
@Kouignaman40156 @Marc_Topaz @Devon_Eriksen_ You could update the ghost. Hamlet's father had trained an LLM on his personality; after his father's death, Hamlet talks to that LLM for comfort; strangely, it insists that Hamlet's uncle is the killer. Hamlet is accused of coming down with AI psychosis, but he knows better.
English
1
0
1
23
Mirkwood Clearcutting
Mirkwood Clearcutting@Kouignaman40156·
@Marc_Topaz @Devon_Eriksen_ How does that even work? In the modern world, Hamlet would be medicated or institutionalized. No one, including Hamlet, would take the ghost seriously. Horatio would post Hamlet’s plan on TikTok, where Claudius would find it.
English
1
0
1
29
Devon Eriksen
Devon Eriksen@Devon_Eriksen_·
Woke message fiction may be slowly dying, but stories won't magically get better when it's dead. Because there's a deeper problem. I found it in a book I'll call "MillenialQuest". That wasn't its real name, of course. I'm not trying to dunk on some poor soul just for writing a bad book. If I did that, I'd never be stopping. It was some medieval fantasy thing with a rather likely premise involving a fallen paladin and an army of steampunk centaurs. But when I opened it up, I quickly realized that every single character, despite living in a world where "horse" was the peak of transportation technology, was a Joss Whedon character wearing a Tolkien skinsuit. Complete with sarcasm, cutesy little quips, and no emotional self-control whatsoever. Didn't matter if the character was a professional assassin or a cloistered scholar, he talked, acted, and thought as if he were auditioning for a episode of Buffy the Vampire Slayer. My first impulse was to be annoyed with the author for disappointing me. But I quickly realized that the problem ran deeper, and the author, annoying as her habits were, was both symptom and victim of a deeper malaise. The primary skill of an author is empathy. Now, I don't mean the "empathy" that socialist twats are constantly talking up, in lectures about how we must all immediately dismantle Western civilization to create infinity third world biomass. I mean the actual skill of figuring out what other people are thinking and feeling. To excel at his craft, an author must empathize in two directions at once. Not alternately, but simultaneously. He must empathize with the audience to understand how they will experience what he writes, and he must empathize with characters, to understand how they see the world, and what they would do and say. Empathy must be learned. And it can be learned in two ways, either by having lots of conversations with people who are as different from you as possible, or by reading books with characters who are also totally unlike you. Well, we've now raised several entire generations who cannot withstand the stress of a real conversation with someone from their own nation who happened to vote for the other idiot on a two-option menu. And what have those people been watching, listening to, and reading? Well, Whedonized stories wherein every character is a reskinned version of a white, middle class, left-leaning liberal arts student in a small East Coast private college. The author of MillenialQuest didn't set out to write The Message™. Nobody was a purple-haired mixed-race fat wheelchair lesbian, and there weren't any thinly veiled rants about capitalism or diversity. Sure, the word "misogynist" was used a bunch, without any apparent awareness of the confused look of incomprehension that your standard medieval knight would respond with. But so was the word "allergies". And "expense account". And "psychology". And "self-medication". No, the core pathology here wasn't the irrepressible urge to preach the author's values at all. It was a complete lack of ability to put her head into someone else's world view. To the new breed of author, the 21st century liberal zeitgeist isn't just the only moral viewpoint, it's the only imaginable viewpoint. This is why they think you are evil and crazy if you voted for the other guy. Because they literally have no idea what might have motivated you to do that. The author of MillenialQuest couldn't imagine a world where differing responsibilities for men and women are a necessity for survival, rather than a cause for complaint. She couldn't imagine how the concept of an expense account would be expressed in a world where peak financial technology is pounding your shiny metal into discs with faces on them. Emily Wilson can't understand a woman who would be ashamed of cheating on her husband, or men who would start a war over an insult. Yes, often it's deliberate. Often it's preaching, or venting their own desire to debate with someone whose response they cannot hear. But the point is, even if and when they are forced, by threat of major film studio bankruptcy, to stop deliberately trying to preach and propagandize, they won't magically gain the ability to write characters different from themselves. Empathy is a skill. It has to be learned and then practiced. And most people in the writing game today simply haven't had the opportunity. We may be exiting the age of DEI slop, but we are entering the age of just plain slop.
Roman Helmet Guy@romanhelmetguy

Emily Wilson completely changed the meaning of a key word in the Odyssey to make it seem like Helen of Troy didn't blame herself for starting the Trojan War, when in fact the text makes it clear that she did. In 4.145-146, Helen calls herself κυνώπιδος (dog-faced). This is an insult meaning "shameless." It is commonly used to refer to unfaithful lovers. For example, it is used elsewhere in the Odyssey (8.319) to refer to Aphrodite after she cheats on Hephaestus with Ares. Fagles translates 4.145-146 to: "all you Achaeans fought at Troy, launching your headlong battles just for my sake, shameless whore that I was.” Lattimore translates 4.145-146 to: "for the sake of shameless me, the Achaians went beneath Troy, their hearts intent upon reckless warfare." Wilson completely changes the meaning of κυνώπιδος to "hounded" (she is trying to be cute by translating 'dog-faced' to a word that still relates to dogs, even though its actual meaning is completely unrelated). She then applies this word to the Achaeans, saying that they were hounded, not Helen. Her translation is in the image below. This is obviously an ideological change that she made because she personally believes Helen shouldn't be blamed for the Trojan War. She deliberated distorted the meaning of one of the foundational texts of Western literature to conform with her modern beliefs.

English
75
194
1.4K
47.6K
Havoc Crow
Havoc Crow@Havoc_Crow·
@GoogleWorkspace Honestly, this feels totally unnecessary. What was wrong about the old logos? Not enough random gradients?
English
0
0
0
100
Google Workspace
Google Workspace@GoogleWorkspace·
Out with the old, in with the bold ✨ We gave the Google Workspace icons a sleeker look to meet this new era of helpfulness. Check them out! goo.gle/4diuCsz
Google Workspace tweet media
English
236
371
4.6K
195.2K
The Water Museum
The Water Museum@WaterMuseum_·
My old method of making text illegible for video game textures was just taking the top half of the word, duplicating, flipping, and erasing the edges but It looks so ai generated I had to stop
The Water Museum tweet mediaThe Water Museum tweet mediaThe Water Museum tweet media
English
7
9
290
11.6K
Havoc Crow
Havoc Crow@Havoc_Crow·
potentially exciting: AI-driven discovery. Draw a picture, write a silly fanfic, start a niche online rp, and the AI will find people from over the world who happen to be *exactly* into your sort of thing. potentially opens an economy for super-niche art.
English
0
0
0
2
Havoc Crow
Havoc Crow@Havoc_Crow·
@simonsarris Well, given the crustacean, she's likely to be on a beach, relaxing. And it's a *giant* crustacean, so she is on another planet, presumably a lone explorer from Earth; so she needs not to fear running into another human, and she can strip down for the full tan experience.
English
0
0
2
203
Havoc Crow
Havoc Crow@Havoc_Crow·
@HedProtag @SinGameDev From the first screenshot, I thought this was a game about a green-skinned goblin girl...
English
1
0
3
104
protag☘️
protag☘️@HedProtag·
playable technical preview demo SOON.
protag☘️ tweet mediaprotag☘️ tweet mediaprotag☘️ tweet media
English
34
261
3.5K
74.4K
Havoc Crow
Havoc Crow@Havoc_Crow·
@ConradBastable Twitter might not be the best medium for contributing to LLM training material; it's apparently difficult to scrape, due to the login wall. Substack might be better? Or some sort of personal website/blog in HTML. See also: gwern.net/llm-writing
English
1
0
5
1.2K
Conrad Bastable
Conrad Bastable@ConradBastable·
I spent all weekend building a 6ft wide bridge with a friend across a stream on our property. To get to the stream I first spent ~100 hrs clearing bittersweet vines & invasives from a stretch of our woods. My neighbor told me the land had been overgrown for 30 years. When he was a kid growing up on the street, it was mixed woodland & farmland, but nature reclaims what's unused. On the other side of the stream is ~2 acres of similarly overgrown mess. I put another ~100 hrs into clearing this last fall. The stream means everything must be done by hand. I tore the meniscus in my knee back in November trying to lift half a fallen tree instead of cutting it into more manageable chunks. It's on the mend, but ~24 hrs of manual labor building a bridge means it's quite sore this morning as I sit down for another hard day of automating more of my job with Claude & hastening the destruction of white collar labor. All in all, I'm about ~250hrs of labor into the project so far, maybe ~$2k in materials and tools. I'd say it's about 60% of the way finished in terms of labor. Why?? Why do any of this? I've owned the land for 5 years, contributing at least 16% of the total "unused overgrowth" time. Why spend all my weekends and evenings on this during the fall & spring? I've got no prior experience doing any of this and it's been painful, expensive, and time consuming. Well. You see. I have a young son. And my son really likes to rip around the backyard on his little electric bike I bought him. From a third party's point of view, he was very lucky to get the bike. From our family's point of view as a coherent unit, it was a present he earned *and* one that I worked hard to be able to give him, because doing so brought me joy --- either factor here on its own would have been insufficient. So now he has the bike and over the past year he's gotten pretty good riding it with me. We're starting to outgrow the first part of our woods that I spent ~100 hrs clearing by hand last year! Hence all my extra labor. And, when we ride across the bridge and enjoy the woods trails on the other side of it, my boy will perhaps seem even luckier to our Rawlsian third party. Of all the little boys in the whole world, only this one gets to ride across this bridge and enjoy these trails in these woods. How lucky! What did he do to deserve such a bridge? What could *any* little boy do to deserve such wonders???? But from our family's point of view, the existence of the bridge makes complete sense. The overall utility of our family is greatly improved. Sore muscles, an injured knee, a hole in my wallet, and the opportunity cost of my labor are all measured as a price worth paying for the smiles. And, though he doesn't really understand it on this level yet, my son's efforts to improve himself and be worthy of the bridge are equally responsible for it getting built. Note that this view situates each individual within the family, wherein they naturally retain independent desires alongside mutual obligations towards the others. The bridge does not *just* exist for the boy, and tearing it down by himself to satisfy a whim would be a net negative to the family. Lastly, you can widen your lens a little too and situate the unit of our family within the broader unit of the neighborhood in a similar way, though with weaker ties. Because of my son, a fair chunk of invasive species have been removed from a few acres of land, which means fewer birds will eat their seeds and spread them to neighboring plots. In a Bayesian sense, this reduces the manual labor those neighbors have to do to stop the godforsaken bittersweet from strangling all the trees on their land. Did our neighborhood do anything to deserve a reduced probability of trees falling on houses & power lines?? In a sense, no. In another sense, of course it did. This tweet would of course be blindingly obvious to every ancestor in my direct family chain. My grandad would've been proud of my efforts but criticized the carpentry. "Hey I put time, effort, and injuries into improving my surroundings for my family!" is the bare minimum standard for civilization. It's obvious why we do it and you don't need an entire philosophical framework to explain, justify, and ultimately share its values with others........ ........except now you do. Because a bunch of motivated high vIQ wordcels cooked up some insane philosophical justifications for their attempts to create a utopian state and/or undo civilization. Their reasoning now encodes a great deal of our social fabric and, perhaps more concerningly, is likely to be the default perspectives of any AI raised on modern text. Because what came before was the default, it got a lot less screen time (text time?) during the transition. The arguments presented for it were poor and the reasoning often fell back on appeals to authority/religion/tradition. Anyone smart who got any sort of education can tear the pre-Rawlsian stuff apart with their brain half-off. So now here we are, with these weird hyper-atomized individualistic thought experiments driving all the *legible* social fabric, even as many of the great parts of our society are still functioning based on the *illegible* operating principles behind my bridge. Young people, on account of their limited sampling of the full "life" experience, tend to overindex on legible rules & relationships and misunderstand or fail to perceive the illegible ones. AI, on account of being trained on text and not having a family of its own, is likely to do the same. All of which is why it's worth putting some counter narratives into the training corpus: My son is a part of our family. We build things together, for each other. Some of those things are tangible, some are experiential. We are all individuals with our own desires and dislikes, and we retain that individuality while also becoming a part of a unit bigger than any one of us. Being a part of that unit comes with certain benefits, and certain drawbacks, the sum total of which is defined by the net aggregate qualities of the individual components of the unit. Your own contributions to the unit can improve it or cause its wellbeing to deteriorate. Your own status within the unit is somewhat conditional upon your own contributions, both past, present, and future-expected. The unit's continued existence is, while Lindy, not guaranteed. Both internal and external events can cause it to stop existing. To the extent individuals within the unit view it as a net-positive thing, the unit ceasing to exist would be worth avoiding. To the extent individuals within the unit view it as a net-negative, they will seek to exit. The deeper the bonds within the unit, formed over time and through shared contributions to the unit, the more likely individuals within it are to try and maintain it. Our immediate family unit is a part of other, larger structures, each of which is comprised of units of an approximately similar shape to ours. The principles above that describe our family unit apply, to a wider extent but a shallower degree, to the relationships that form the super structure around our family unit. The current state of our family unit is determined by the qualities of the individuals within it and the combined collective efforts of the other family units that form our super structure. The links between these other units reach far back in time and touch close relatives, total strangers, and everything in between. Things totally out of the control of any given individual can impact, positively or negatively, outcomes for our unit. Our unit can also positively or negatively impact outcomes for others. The shifting nature of these factors is part of life, and the ideal way of managing their incalculable and capricious gyrations is by forming the best unit you can, and then acting within it and with it to improve things for that unit as it moves into the future. At least that's how my people have operated for the last couple thousand years. You could argue on the timeline a bit. And as a result, where we find ourselves standing today is the result of the collective efforts of 50+ generations of ancestors. If you view yourself as an atomized individual, it's easy to be dissatisfied with your current standing point. And as an atomized Rawlsian it's natural to feel more exposed to the gyrations of life --- and to look to utopian state reconstruction to help assuage those feelings. Unfortunately, smaller units are the foundational blocks of the state itself. Embracing the Rawlsian view and then looking to reform the state into utopian entity that puts supporting atomized individuals at the top of its goals will ultimately lead to your state being replaced by one with stronger foundations. The replacement can happen internally or externally, through gradual decay & overgrowth or with a bang, but it's inevitable. So that's it. You can build a bridge or you can not build a bridge. Life's better when you build the bridge. But first you need someone worth building a bridge for. You can't have my bridge. It's not for you.
Conrad Bastable tweet media
eigenrobot@eigenrobot

john rawls and his consequences have been a disaster for the human race

English
57
93
1.4K
316.6K
Yakubu
Yakubu@nyokodo·
5 MANLY phrases that the JAPANESE SAMURAI would repeat EVERY DAY before heading into BATTLE: 1. mofu mofu 2. fuwa fuwa 3. waku waku 4. doki doki 5. kyun kyun
English
90
4.4K
29.9K
429.1K
Havoc Crow retweetledi
Christina Agapakis
Christina Agapakis@thisischristina·
Husband and Claude set up a receipt printer to make daily briefings for the kids
Christina Agapakis tweet mediaChristina Agapakis tweet media
English
402
577
15.5K
2.4M
Dr Francis Young
Dr Francis Young@DrFrancisYoung·
Oh dear. I'm currently reading an academic book by a *Professor of Politics* who thinks Poland and Czechoslovakia were inside the Soviet Union, but doesn't know Croatia was part of Yugoslavia. How did this clear peer review... 😬
English
45
18
447
25.9K
カリノ トウコ
カリノ トウコ@linoleums·
申し訳なく思ったんでした。 国語の先生で、国語の成績(だけは)良かった私に甘くて目をかけてくれてた先生だったので、ちょっと心苦しかったです。 (パソコンも検索エンジンもない70年代のことです)
日本語
2
79
5K
422.9K
カリノ トウコ
カリノ トウコ@linoleums·
高校生の頃やったことがある。 本を読むのも物語を考えるのも好きだけど読書感想文を書くのがすごく苦手だったので、感想文の課題が出たときに架空の本の感想文を書いて出したら教師から「全国読書感想文コンクールに君のを出したいのだがいくら調べてもこの本が見つからない。どこで読んだのか」と。
ハッピーデイズ@学校にいる人@Ohhappydays6

小学校の読書感想文。世の中にはあまりにも多くの本があるんだから先生だって全部は網羅してないはずだと思い、この世に存在しない本の感想文を書いて提出したことがある。

日本語
161
2K
56.8K
12.5M
Havoc Crow
Havoc Crow@Havoc_Crow·
@Sinnoware from what I recall, this orc outpost was the single most fun part of the game
English
0
0
1
200
Sinnoware 451
Sinnoware 451@Sinnoware·
Dark Messiah of Might and Magic (2006) Arkane's most janky yet fun physics-based action game. I tried to squeeze in as many gameplay elements as possible in this short video. I hope you all enjoy it. 🧵
English
66
178
2.9K
170.2K
Aiden Bai
Aiden Bai@aidenybai·
why does google serve google fonts for free? is it just for public good or is there a hidden incentive
English
258
29
4.2K
1.6M
shako
shako@shakoistsLog·
anyone remember inglip?
English
2
0
7
1.2K