In the “Machine Learning for Art” scene there is “before GAN” and “after GAN” (2015 > now). The vast variety of methods & ideas of the early period, got homogenized into a endless stream of GANerated pixels, that tether on nihilism and yet are sold as magic. Artificial Progress?
The Wondering Earth was a box office hit during the spring festival in China. And Chinese showing their ticket , printed Only Communist Party Can Save The Earth. No I am not kidding @BaldingsWorld@Jkylebasspic.twitter.com/KjO3EXmcpW
‘Seek out the faceted moments of geomediation at the Hermits Rest, after staring down The Abyss. Layers upon layers, Bright Angel Shale, Vishnu Schist and Zoroaster Granite. The Great Unconformity (an absence, a non-layer).’ https://t.co/KOEP6nantK
You know what’s the problem with Singlish’s street cred? Every joker who has ever heard it before thinks he or she can speak it. Just ask that angmo who’s been in Singapore for a few weeks, and you may hear him or her boast: “You think I cannot speak lah?” Actually, angmo, I dun think hor, I
know – because that use of “lah” is so salah.
Poor, poor “lah”! It has kena so much abuse that sometimes we wonder whether we’re still a Western colony. Come on, show some respect! “Is ‘lah’ a note to follow ‘soh’?” isn’t funny – we’ve
heardit a gazillion times. Such wilful ignorance isn’t cute. “How go to Orchard Road lah?” “Lah your friend is so beautiful.” “The chicken rice nice-nice leh lah!”
What the fiak. (Yes, that’s how we say it.) “Lah” has to be like among the most abused words in the history of abused words. So let’s get the rules right for the sake of our sanity. “Lah” is firstly used at the end of a sentence or a main clause – nowhere else. No “Lah you is so funny” or “I take lah the bus home”. This isn’t French. But you can say “Dun care him lah, let’s go!” or use it with a filler like “OK”, as in “Please lah, OK?” (By the way, this “OK” doesn’t mean OK; it means “for God’s sake”. Surprise.)
Secondly, there’s no need to use “lah” to end every sentence. Once is enough to set the tone – unless you want to change this tone or to be irritating. So it makes sense when you cry, for example, “Help me lah! Lend me money lah!” Otherwise, dun lah-lah-lah please.
Thirdly, “lah” isn’t exactly meaningless. And it doesn’t mean “dude” or “babe”, and you shouldn’t say “How are you lah?” when you mean “How are you,
bro?” There are even multiple meanings to it, and each meaning is defined by the context in play. I can identify three main types:
The pleading “lah”, as in “Go away lah!” or “Go and die lah!”, and both of these mean “Please get lost”.
The emphatic “lah”, as when you hear “You see lah!” The emphasis here is on you seeing – which presumably you haven’t been – and the line therefore means “Didn’t I tell you to watch out?”
The affirmative “lah”, as in “Steady lah!” or “Solid lah!”, both of which mean “You’re impressive! Keep it up!”
These three “lahs” are also differentiated by tone, and so you need to learn to enunciate right too. None of them involves the sing-song “lah” of
The Sound of Music. The pleading “lah” sounds like a deflating balloon: “laaah”. The emphatic “lah” is a spurt ending on a higher pitch, like when something drops on your foot. The affirmative “lah” is also a spurt, but it pulls downwards after going up, thus showing the very control of feeling it signals.
So, everyone, please lah, OK? Dun simi sai also go lah. Use your “lahs” sparingly and accurately, and you will win over the easily wounded hearts of native Singaporeans faster than HDB upgrading!
– Gwee Li Sui is a poet, a graphic novelist, and a lite-ra-rary critic who also likes to talk cock sing song.
Seriously, if we can’t make up words willy-nilly, or borrow them from random languages and underground emoji codes wtf did we even build the internet for? The whole point of the internet was to act as a complexifier of language 🤬
Prompted by a conversation with @infovore@instagram.com about our separate experiments with timelapses over many years (given the timelapse machine I’m currently building, which is right now taped to the win…https://t.co/xDYlf9dkzgpic.twitter.com/95McEF6s6Q
Word of the day: “slipshape” - in contrast to “shipshape” (tidy, ordered, regulated), that which is “slipshape” is characterised by fluidity, uncertainty, shape-shift & flow.
Coined by Alice Oswald in her book-length river-poem Dart, “a songline from the source to the sea”. pic.twitter.com/fH0cNo2K2U
New research finds Australia is installing renewable energy faster than any other country, a trend that will allow Australia to meet its economy-wide Paris targets five years ahead of schedule if politics doesn’t derail the trend. https://t.co/pFanmeaTIo
Senior EU officials have told me bluntly they know Brexit was won through illegality and cheating by Leave. Leaving the EU is not the ‘will of the people’. Brexit is the will of a British government complicit in ignoring data crime, electoral crime and Russian interference. https://t.co/G1HdqBPCzw
Srinivas: Anthropology cannot be satisfied to train students for a crumbling world, but rather should allow them to imagine a world that could be and then give them the imagination to build a future that only they can see.’
Complexity concept of the day: The greatest challenge today is building collaborations among people who don’t understand each other for collective benefit.
Just learned the amazing word “apophany” from apophenia, the tendency to see patterns when they aren’t there, i.e. faces in clouds, nonexistent patterns on the roulette wheel. So an apophany is like an epiphany except you’re wrong.
Within the past 24 hours, France, Russia, & the US have all conducted test launches of unarmed nuclear missiles.
France: 1 missile launch from a warplane
Russia: 1 ICBM Yars from Plesetsk Cosmodrome
US: 1 Minuteman III from Vandenberg AFB
This is a picture of the far side of the Moon and Earth beyond, captured on Feb. 4 by a camera on board the student-built Chinese DSLWP-B/Longjiang-2 satellite and received by the amateur-operated Dwingeloo Telescope in The Netherlands. (HT @AJ_FI@radiotelescoop) pic.twitter.com/pQAJaszEH3
Here’s @tobias_revell and I’s most recent work ‘Augury’, an installation that parodies the faith in algorithmic prediction by combining machine learning technologies to create a new form of divination based on the flight patterns of planes: https://t.co/Lg8OQXXuXl
The first ever simulated image of a black hole, by astrophysicist Jean-Pierre Luminet (1978). The surrounding accretion disk was calculated on punched cards, then drawn dot-by-dot “directly on negative Canson paper with black India ink.” Complete story: https://t.co/EF7ekbeIBVpic.twitter.com/okDfmIa4XS
“Magic is insidious, and therein lies its danger… In magical rites the inversion of letters serves the diabolical purpose of turning the divine order into an infernal disorder.”
For professional reasons, got a new phone & been testing the Google Assistant. Out-of-the-box, if you say “good morning” it will start to read Reuters & CNN news. This is the exact opposite of what I perceive as a “good morning”. Overall, it feels designed for a alien race.
I admire @GretaThunberg, I respect her, I find her courage inspirational, and I also recognize that having these feelings won’t change the political inequalities, structural economic and social inertia, and climate feedback dynamics that are going to kill us all.
US and UK bring to mind the words of Panait Istrati when visiting the 1930s USSR: “All right, I can see the broken eggs,” he said. “Where is this omelette of yours?” @mrjamesob
Forest supercomputers, computational landscape architecture, and WiFi’s passage through trees—“landscapes sown specifically for their electromagnetic-propagation effects.” https://t.co/arQVqK2XGWpic.twitter.com/GfZubqLk8w
’[E]veryone has a religion, whether admitted or not, because it is impossible to be human without having some basic assumptions (or intuitions) about existence and the good life.’
“A huge Chinese middle class that is thought to number 400 million has brought about a fast-moving extinction crisis for many animal species.” https://t.co/tOMsPrDQmp
Three leaves from a Tibetan musical score used in Buddhist monastic ritual with the notation for voice, drums, trumpets, horns and cymbals.
From the collection of the U.S. Library of Congress pic.twitter.com/TFWXQfbbab
How is our business preparing for pointless brexit? Cancelled travel in April, applied for export licence, stocked up on materials in studio, have the FoAM international network as backup for contracts in EU & loving the sudden horrified awareness of invisible infrastructures.
The crossword are probably the only mainstream puzzle format that requires a full-blown postrational civilizational OODA loop running in your head. You can solve sudokus with pure rationality, but crosswords require you to understand both the zeitgeist, and others’ maps of it.
I revisited the Museo Xular Solar. If Buenos Aires *is* a labyrinth then its Asterion is not the obelisk but this museum – a labyrinth inside a labyrinth with its own universal language and music, its own form of total syncretism, its own Glasperlenspiel. #PanChess#XularSolarpic.twitter.com/91Kee663ZV
It’s impossible to predict the future. Any models or scenarios will inevitably be mostly wrong. So I recently described the goal of the Future Committee as “being wrong about the future in the most useful way possible”. #futurism
“Memory’s truth, because memory has its own special kind.
It selects, eliminates, alters, exaggerates, minimizes, glorifies, and vilifies also; but in the end it creates its own reality, its heterogeneous but usually coherent version of events.
I’ve written before about BigGAN, an image-generating neural net that Google trained recently. It generates its best images for each of the 1,000 different categories in the standard ImageNet dataset, from goldfish to planetarium to toilet tissue. And the images it produces are both beautifully textured and deeply weird. Some of the categories - scabbard, rocking chair, stopwatch - are delightfully aesthetic.
[scabbard, rocking chair, stopwatch]
Google has made the trained BigGAN model available to the research/art community, which is nice, since people have estimatedthat today it would take around $60k in cloud computing time to train one’s own.
But there’s more lurking in the BigGAN model besides the 1,000 ImageNet categories. The model thinks of each category as a big set of numbers that describes exactly how to smoosh and stretch and color random noise. Following one set of numbers will transform noise into a flower, while following another set will turn that same noise into a dog instead. But another thing a set of number is, is a position in space: latitude and longitude for example, or x,y,z coordinates - in math terms, we call the set of numbers a vector. And in machine learning, all the positions in space (granted, an approximately 100-dimensional space) that a model’s vectors can point to is called vector space.
So one set of numbers - the flower vector - points you to some location in vector space, and another set of numbers - the dog vector - points you to a different location.
[daisy, saluki dog]
But here is where it gets fun. The vectors are just numbers, which means you could, in theory, average them. What happens when you average together “saluki dog” and “daisy”? There’s no ImageNet category there, so what’s lurking in that spot in vector space, halfway between the two? Delightfully, dogflowers.
This, it turns out, is so cool. Joel Simon has put together an app called ganbreeder.app that lets you mix and match categories.
He is undoubtedly one of the most influential people in human history (that you have never heard of).
As you and I may well be alive today because of his work, here’s a (short) thread about him on the 76th anniversary of his death. pic.twitter.com/6Dx31bYWR0
just biked from Taipei to New Taipei City while listening to the entirety of DJ Shadow’s “Endtroducing…..”, extremely into it pic.twitter.com/6MqyQRBZo8
clouds workshop day two, wonderfully rich offerings: desert poetry and balloon expeditions, contrail art, ice-cream artillery, lightning guns. the #intheclouds hashtag has a snapshot.
“Adults keep saying: ‘We owe it to the young people to give them hope.’ But I don’t want your hope. I don’t want you to be hopeful. I want you to panic. I want you to feel the fear I feel every day. And then I want you to act.” https://t.co/fnvErGbfOI
Ten years ago I set up a small DIY label for noise & experimental music. After 36 releases by artists from Italy, China, UK, Australia, Hong Kong & Germany, we’re closing shop & everything is on Bandcamp for free download. A few takeaways: https://t.co/cFLegqLuTt
How can you do great AI research when you don’t have access to google-scale compute? By being weird. The big tech companies are obsessed with staying nimble despite being big, and some succeed to some extent. But they can’t afford to be as weird as a lone looney professor.
“Santa” is very important because it teaches kids that adults are capable of staging large scale, leaderless, and decentralized deceptions just for the hell of it
Um yeah sure ok, Elon Musk can shoot his car into space but my idea of putting authentic mastodon bones into oversized space suits and placing them on the Moon next to a homemade wooden rocket in order to confuse future archeologists is “unreasonable” and “a waste of resources”
as a composer I am applying into the vast and limitless void here, for the next women led feminist intersectional interspecific cultural luxemburgist sci-fi #scifi hmu DMs open