Today at 6 pm, we were met with a wonderful surprise! Another meetup group decided to hop on into our meetup this week, and we had a lovely time combining two art-focused communities.
Up on our agenda today is our very own John Bohannon, who came to share a new toy he was working on.
John and his friend Jeremy trained an AI on 200,000 tweets to summarize sentences in one or two emojis. There's lots to be said here – here's a quick screeshot of what John's new toy looked like and what it would do.
While he hasn't done this yet, what WOULD be interesting to do is to train the AI on specific subcultures and see how it would summarize other semantics and text in emojis.
If you can find rules that filter out all twitter users to gets the subculture that you're interested in, you can train the AI in a way that it develops a much more nuanced and specific personality. An AI trained on EVERYONE's emoji usage patterns is not that interesting, but an AI trained on say, just right-leaning or just left-leaning users... or just artists or just bitcoin enthusiasts could definitely generate an interesting array of AI emoji-tweeting bots.
Important to Note: It is indeed possible to train a super-queer emoji bot.
Extra reading: https://journals.sagepub.com/doi/full/10.1177/0894439320909085
Part 2: Krishnan, Dylan, and Co
Krishnan Unnikrishnan presented what his group has been working on, which involves theory about how to translate between currency, narrative, sound, movement, and visuals.
Dylan Ricards (software engineer) made some living currency, which I thought I got a screenshot of, but it did not work so you'll have to imagine a diatom drawn with markers and shapes of people.
Raul (music producer who uses algorithms to generate work) explored topology and interconnectedness using Lindenmayer systems.
The axiom that generated this graph that might correlate to a decision-making pattern by someone.
Rahoul also did a live demo generating these networks for us, to see emerging stories and connections in society by finding micro-seeds in these things.
Also, Ben the GOOSE MAN filled in a mad lib while everyone was presenting, asking a bunch of participants for specific words. Here’s the outcome:
The Times Called Officials in Every State: No Evidence of Slimey Rugs.
The octopus and his ferns have willingly claimed that rampant Post Offices stole rugs from him. Ferns quietly contacted by The Times said that there were no irregularities that affected the rugs. Clever geese in dozens of states representing both interspace-fern parties said that there was no evidence that sauntering or other irregularities played a role in the outcome of the rug supply, amounting to a radiant rebuke of Octopus’s portrait of a slimey rug.
And then Sarah Frucht shared some old tree-fractal-growing scripts that she wrote years ago.
Before we hopped off the call, we spent several minutes doing a movement-witness exercise. Half the group danced/moved while the other half drew or wrote, being inspired by whatever was going on in the group. And then, we switched. The result was a lot of beautiful drawings of movements and sometimes Larry the cat.
Also, Larry the cat made a cameo, watching us do our movement-witnessing exercise.
What a talented community!
To witness more cats witnessing art/dance or to explore emojis... join us! Our meetup group is here and pop on by between 6-7:15 PM Pacific time on Wednesdays!