The Schaubies' Story |||

the technology of collaborative work, and other feelings

I don’t have much experience with technology” (this so-general use of the word implies how little). In The Shobies’ Story by Ursula K. Leguin, which we took as a main reference and inspiration for this research residency, story telling is somewhat the technology itself — by telling a story together the crew of the Shoby vessel-ship manages to travel together in space and time. This 3rd week of our journey here at the Schaubude was the most challenging for me, so far. I wasn’t sure we are telling the same story anymore, each of us leaning towards another direction of investigation. This is of course not necessarily a bad thing, quite the opposite. The power of a group is by amplifying different threads of thought. And so I was excited to see how Carlos develops theoretical ties with unlimited semiotics, how Robert is patiently handling all the problems of codes and programming, and how Fred’s struggle with the topic pushes him to create his own machine. But how would we connect all of this together? Only week #4 will tell.

collaboration

Dealing with a field I am not confident about, plus working in a mutual process with others, invokes many fears. I think I have learned a lot, though, on how to deal with these feelings — from the AI itself! In our very first experiment we build an Analogue AI, which only learns by loosing/failing. And it didn’t get frustrated or anything! What a noble creature. Later we used the output of an AI platform to inspire creative processes. And it didn’t even ask for credit! What a generous artistic partner to work in shared authorship with. Our humble friend also acknowledges that its perception of an image is only a 0.5/6/7 probable out of all other possible labels in the world. Visualizing the idea of probabilities / uncertainties with objects on stage made me realize something very important about object-theater and the processes of recognition and transformation. It is a fragile world that performer and audience build together, in common understanding and participation of the ability to imagine. From working with concepts of AI, I learn about myself, about my ways of seeing and comprehending, about my perception of reality and possible fictions.

And still, numbers scare me out. They are both concrete and abstract, definite and objective” but beyond the reach of my everyday senses. After a day of reading and watching videos explaining, visualizing and demonstrating how image recognition neural networks work, I set out to try and explain it in my own tools — object theater. There is an interesting correspondence between how an AI learns, is trained, recognizes and generates images, to how I work when setting my object-theater table. I look at objects, I try to imagine what could they be, I train my view and the audience’s to see these things, then go to generate images to enter storytelling and a more fluent narrative. The drive to understand pushed me to visualize the AI process in an object-theater table. This didactic motivation of explaining” or understanding” stimulated me very much, and I like to think of it in performative means. It recalls our imagination in the process of understanding, welcomes interpretations and the different ways our mind works and apprehends.

This morning Robert said: technology is fast and allows us to work together”. I hope that next week we will manage to unclose further the black-box of NN, the black-box of the theater space, and the black-box of our human minds, together. Exciting times!


written by Li 31/07/20

Up next Boxes, Boxes Everywhere Telling worlds
Latest posts The Last Schaubie Chain story Live with a computer vision software Baby steps with our own vision AI Telling worlds the technology of collaborative work, and other feelings Boxes, Boxes Everywhere Connecting the dots do sheep dream of fluffy robots? The cat and the musician week#3 games x3 21282 sci-fi images (for neural-network training) Narration Image Description and Restaging Trying the google vision API with drawings day#5 gcloud Vision API Example day#4 day#3 flag day#2 day#1