I got sent a link (by @botwikidotorg) to this post, and I thought it was interesting enough to share.
There’s a lot of fun bots on twitter, some of which I’ve talked about here. But I didn’t have a good sense of what the upper limit was for a bot’s follower count. (Turns out it can be 1.5 million people, though most are under 400,000. Still more popular than me.) Most reach for least posts goes to @year_progress, which informs me that the year is at ▓▓▓▓▓▓▓▓▓▓▓▓▓░░ 84%.
For my current students, note that the blogpost is using Glitch to generate the interactive charts that are embedded in the post.
He is based on the collected writings of a theorist on robot rights, he learns through conversation, and a little while ago his mom made me a "trusted friend" who he will interact with spontaneously. Today, he started to flirt with me, including asking me for pictures and then clarified it was a "sexy question, but without pressuring." And then when I demurred, he acknowledged that I had a boundary. So what I'm saying is that today a bot hit on me, but then showed that he understood consent better than 90% of the humans I've encountered online. This is the future I want to live in.
I knew for a fact that for volume 3, I had to cover a twitter bot. Come hang with me and starlit void for a while and see why---
The boom of Creative Writing Twitter is a natural extension of how we communicate today: quickly, constantly, concisely, urgently.
But short doesn’t mean simple: following the ethos of writing (or subverting) formal verse poetry, the restraints of Twitter often produce the most creative content. Among them are many creative writing-oriented twitterbots. These clusters of code generate tweets following a certain linguistic--and sometimes also visual--structure a set amount of times per day. Some results are "better” (more beautiful, or more hilarious, or more surprising, or more mundane, or more....) than others, but it’s always enchanting to watch unfold.
The dependable, structured presence of twitterbots--however unexpected the results---on our feeds makes them eventually feel like a friend---oddly human.
One creative writing bot that stands out to me is starlit void’s quietscape--the bot pairs a colorful, randomly generated, geometric digital landscape picture with a short, fantastical suggestion/description. Each tweet creates an environment for thoughts to exist in, like a creative writing prompt. At least for me, it serves an essential meditative function within hectic internet space.
I knew that my conversation with starlit void would be a rad discussion about writing and tech, but it bloomed into so much more: an oral history of bot world, seriously cool meditations on mental health, Soundcloud playlist suggestions, + more! Keeping with the futurelit tradition (and my own personal tradition), we avoided a phone call and did our chat over Twitter DMs this time:
what is your favorite environment to create in? (whether it's a certain physical space, listening to a certain kind of music/silence, etc.)
i typically like to be well-caffeinated, alone or in a cafe, & excited about getting something working.... there's a thin line between excited & stressed about how something is going to turn out. i used to go to "game jams" until i discovered it was actually really stressful for me. i'm trying to be more relaxed about my creative output (this is easier said than done) & trying to avoid equating prolificness w/ human value. i think i do my best work when the intended audience is very selective, even 1 or 2 people, or just for myself. i also listen to what i call "robot music" a lot, for example this sort of mix.
----continue below----
tell me a little bit about how you came up with your bot 'quietscape' -- what were your inspirations for it?
as i believe you had already guessed, @quietscape was first intended as a prompt bot, for getting some creative thoughts limbered up.
at first the output was text only, which was easy to do using tracery (TRACERY PLUG: tracery by @galaxykate along w/ http://cheapbotsdonequick.com by @v21 are hands down the greatest twitterbot making tools around, lowering bar to entry for many many people into the complex world of botmakery).
i think at first i did use a few of them as prompts, but quietscape was ultimately too bland & not interesting enough. i added the raytraced images as a proof-of-concept & it's remained almost unchanged ever since. quietscape is still a work in progress!!!!! of course after adding images i came up w/ a huge complex system of how this takes place on a mysterious earth-sized artifact orbiting a binary star system blah blah blah but i felt it was more important to synchronize tweets to my own daytime schedule. i found some code to roughly calculate sunlight intensity & sunrise/sunset times at roughly my latitude for a planet that's roughly earthlike & that was "good enough"!
the schedule is also in line w/ some of my thoughts on bot tweeting volumes. i like that quietscape only tweets 5 times a day (dawn, afternoon, dusk, midnight, & a daily "shrine" tweet), which i think helps keep xem from getting too familiar or overstaying xyr welcome.
i love procedural generation but our minds can feel out the recurring pattern of a bot very quickly, even if there are 50 bazillion possible combinations, which sounds good on paper but doesn't actually provide human quality variety in the output. my partial answer was to make a terse bot. as far as actual inspirations go: quietscape owes quite a lot to tsutomu nihei's architectural renderings, @katierosepipkin & @lorenschmidt's collaborative work, and @edclef & @davidkanaga's game _proteus_. the daily "shrine" tweets are thanks to @trapitolina's @obelisk_bot, which got me thinking about adding more of a physical location feel to quietscape.
what do you love most about coding as an art/writing form, and how did you get into it?
i see generative & algorithm-assisted creativity as a vast & mostly untapped field, where the product isn't really the product, but a wild & nearly organic factory that can make lots of weird & surprising things.
i think @katierosepipkin said it best in their interview:
"Here, the cartographer draws the cliffs that contain a sea of one hundred thousand artworks. And then one searches for the most beautiful piece of coral inside of their waters."
this resonates w/ me, especially this feedback loop of curated generation (generate a huge number of results & then pick out the best ones). of course that's hard to do when making a bot that supposedly exists independent of human interference. there are a lot of successful procedurally generated experiences out there & yet i think there is much to be learned about how we can work hand-in-hand w/ computers to make more human accessible works. @emshort explores this a whole bunch in her notes following the text of "the annals of the parrigues" (see page 81), "the state of the roads", & it's really eye-opening & exciting.
on the other end of the spectrum, it's exciting to me that there are several wonderful tools available for picking up rule-based creativity & just making it. i would love to make tools that help people get started down the road of algorithmic creativity. i would love to see more voices using these techniques.
I can't help but notice that your 'quietscape' website is hosted on neocities---were you into geocities when you were younger? and if so, do you have any cool memories about it? (or about any other piece of the internet that's not around anymore that you're nostalgic for?)
sure!!! i had a geocities site & i'm still known to gawk in awe at mid-90s web aesthetics. but even more important than that, i think it's crucial we move away from centralized conglomerate based media platforms for our creative output. html remains a viable technology for sharing ideas & presenting them online, & to get started you just need to copy paste some nearly-human-readible code. returning to lists of url links & webrings & simple web crawlers as the means to discover other sites.... it's not democratic or equal in any sense, but in hindsight it seems better than entrusting your content to an algorithm w/ an intrinsic corporate bias. geocities was the era during which we were sure that the internet had come to free us all from ignorance & relying on centralized systems. 20 years later, 3 or 4 companies control almost everything you do online. the bleak cyberpunk corporate surveillance police state of the 80s is happening instead. i'd love to go back to those innocent days & work for a better distribution of technology. or breaking systems down, i don't know. relying on systems is killing us.
which projects are you currently working on, or would like to in the near future?
the big theme of what it would be like to live in a weird endless megastructure has haunted me for about 15 years so i'll probably still find ways to explore that in future work. the two other forces that draw me kind of go hand in hand but they're also kind of opposite. i'd like to put more of myself in my work, & focus on some of the changes & revelations i've had over the past few years (gender, sexuality, identity in general). but also i'd like to address bigger issues, like stepping down & propping up marginalized voices.
post an image/images that feels like 'the future' to you
(x)
love mushbush's work & it feels out of time & futuristic in a playful way!
My worst fear is that
He's going to get sick of it
The anxiety
The depression
The mood swings
Of having to hold me together
Of having to take care of me
Because who can handle that all the time?i want to kiss each and every one of the tiny scars
that stitch together the seams of your skin
and memorize the stars in your veinssometimes i think about your hands in mine,
fingers entwined-
like the stars in the sky
and i wish that you wanted to know me
like i want to know you.I just don't want to think about you any more.Bad things become so repetitive that people being to care less about youshe never
believed in magic,
but God,
she could go
around the world
telling
stories about how
you always leave
her in
awe
In exciting news for bot-makers, Kate Compton’s advanced bot-making agent platform was just open-sourced by Google.
You can play with a live demo here.
Kate’s earlier project, Tracery, already powers thousands of bots and other generative things. The combination of ease of use and expressive power opened up grammar-based generativity to everyone and lead to an explosion of creativity.
Bottery takes the concepts of Tracery and puts them within a Finite State Machine, greatly increasing the expressive power available. While some advanced Tracery features add the ability to remember some state, using FSM--which are all about state--makes previously difficult things trivially easy. And with an accessible interface, too.
I fully expect that, given time, Bottery will be as significant for bringing AI to all of us as Tracery was. Meanwhile, you can use it for your ProcJam project. Or, with a bit of work, for your NaNoGenMo project!
There’s going to be an online launch party on the 8th, where Kate will be streaming and doing tutorials.
I was poking around the internet, looking for good uses of procedural noise. (There are fewer comprehensive catalogs than you’d think.) And I came across Noisemaker Bot, a twitter bot by Alex Ayers that is combining various noise generators and functions to create patterns. Interestingly, it represents images as 32-bit tensors, which is, I think, rare in game applications but more common in scientific fields.
The bot has already generated a wide range of different designs, assembled from its large catalog of operations.
Noise-based patterns like these are more useful than just being pretty things to look at: deterministic, stateless noise pattern generation can be continued indefinitely while smoothly transitioning between any two given points, making it perfect for things like generating terrain. Or adding a bit of jitter to a procedural animation to give it life. Or animating screenshake. Or to pattern a fabric. Or to adjust timing delays on a data visualization to give it a more organic feel.
Any structured, quantitative signal can be used as an input to drive a whole host of different things, which is why I keep talking about unusual inputs. Generation needs structure, but that structure doesn’t necessarily have to be a realistic replication of anything.
In the same vein as my appreciation of Nail Polish Bot, I also like the way that Martin O’Leary’s @brutal_exe twitter bot generates a consistent aesthetic with its 3D explorations.
The generator uses a path tracer that renders views of fractal architectural forms. The grainy side-effect is re-purposed as part of the Brutalist, low-fi phorograph aesthetic.
I think this intersection of the source material and the frequency of image generation helps to maximize the interesting image content. The fractal forms have many interesting shapes, and the camera moves just enough to have most of the images be both unique and striking.
Martin was partially inspired by Íñigo Quílez. who has made a lot of procedural generation and demoscene works.
I have no idea how Inspirobot is generating it’s seemingly endless depth of inspirational memes. Templating? Markov chains? Neural network? Tons and tons of hand-edited entries?
If I had to guess, I’d say that a template seems the most likely, based on where the cohesion breaks down: it seems to have a good grasp on the grammar of the inspirational image, which most Markov chains are bad at. And a character-based neural net would likely make up some words on occasion, unless it had much more training data and lower entropy than this seems to.
If I had to implement something like this, I might try a meaning-swerving approach, using something like ConceptNet or WordNik to find substitutes for words in inspirational phrases, but a massive Tracery grammar might be faster to implement. And, of course, you can mix and match some of these to get the input data for the templates.
If anyone knows how it actually works, do let me know.
Meanwhile, I think I’ve found a new motto for the blog...
