Most of this week was figuring out a rough draft of a thesis proposal for this coming Wednesday, planning for a talk series related to the thesis, and figuring out an abstract so I can ask folks to be on my thesis committee.
So many things were dependant on me articulating the thesis ideas. I’m excited about it, but the idea feels like a big slippery fish I’m trying to get my arms around, but it would rather wiggle around than let me clearly see it.
I keep thinking at if only I had pivoted earlier, I could have spent the entire summer sorting it out. I’m trying not to dwell on that because I needed that time to make what I wanted (birds) and realize I could use that as part of a stronger theme. And besides, I bet it’ll be easier to articulate when I’m not stressing to get something written for a deadline.
I realized the winter solstice is now the next sun event, which means the 24-hour Solstice live code event is coming soon, and I still need to learn how to make sounds with SuperCollider! This week, I made the leaf program trigger notes when the leaf ate food. It didn’t sound particularly pretty, but it made sounds!
I do feel like I’m successfully taking some of the coolest courses at MIT. I’m TA’ing for Zach’s Drawing++ course, and getting refreshed on the material is nice. Vision in Art and Neuroscience continues to be fascinating, and thinking about perception, what is hardwired and what is learned (and what of that can be unlearned). One thing is talking about how important edges are, and how when staring at an image of blurry colors, the colors can start to fade. Since a fair amount of my work has gradients and edges, it’s giving me things to think about.
I wrapped up a version of the chimney swift that felt like it was ready to share. I made a new feather tract (a way to draw or locate shapes on a deformed XY plane) for the body. For the body, I draw a center point “spine”, draw a line at an angle off to the outline of the body, and then curve the line connecting those points. The value of 1 is close to the back, -1 is close to the belly, and 0 is at the spine. The other tracts for eyerings, eyebrows, and chin are defined similarly, except that the distance away from the reference curve is based on a global scale instead of the shape.
I wonder if I should map a separate plane to the head, neck, and body separately and then blend them, because I think the angles of the body feathers start to act funny near the head. But getting the body was all I needed to add the lighter portion of the swift. (For example, this absolutely won’t work when we have a goose’s neck, but that will be a problem when I get to the Canada Goose!)
A new challenge is actually having a few too many decisions: how exactly should the different feathers on the swift be lit up? It seems like feathers are a bit translucent against the sky, plus reflect light, so it requires more artistic eye to figure out how exactly to draw things. But I’m also afraid to spend too much time there, because there are still issues with how shapes are drawn on the bird (I couldn’t quite figure out the right equation to correctly shape the eyebrow or the tips of the primary feathers to match reality). My plan is that at some point, the bird system will be pretty solid, and I’ll have learned how I want to draw feathers and capture the birds’ personalities, and I’ll go back through all of the birds I’ve made and spruce them up.
Among the stacks of books I’ve been flipping through for my thesis, I picked up “The Nature and Art of Workmanship” by David Pye. There’s sometimes more prestige in design, but a perfect design can’t save it if the builder does poor work. This had some ideas about handmade vs machine-made that I hadn’t heard described so clearly.
Instead of splitting it into “handmade vs machine-made” or “pre-industrial vs post-industrial”, Pye splits it up into “workmanship of risk”, where things can go wrong, and “workmanship of certainty”, where the process produces many things perfectly. Both have their advantages and can be done with quality or poorly. People sometimes want “workmanship of risk” for its own sake, but don’t judge its quality.
Pye talks about how many of the historical objects we admire are examples of workmanship of risk, created by highly-skilled people, e.g., a piece of precisely carved ivory. But now, with a few exceptions (e.g., handmade clay vessels), machines have basically won on the precision front.
Most of the “certainty” things need to start as being “risk”: He gives the example of a typewriter: to mold the letterforms requires risk. At that point, they’ll certainly produce legible letters. However, whoever is typing is exercising some risk since they could create something completely nonsensical with the typewriter.
I’m trying to slot my craft in, programming. Programming is a funny example, because on one hand, I think Pye would say running code falls into the “certainty” bucket. But writing code feels closer to “risk”.
And then to slot in current events, LLMs are trying to mechanically produce code itself. The code produced today is pretty poorly crafted: it repeats logic and changes styles midway through. I don’t know if the chatbots will be able to code better in the future, and I don’t know how much it matters to the proverbial bottom line if the code could be better if we can rewrite it all in another die roll.
Pilgrim at Tinker Creek is still my bedtime book, and I’m nearing the end. I also finished the How to Do Nothing audiobook and started Braiding Sweetgrass.
I also performed visuals with Justin Looper at SomArt Space @ the Hive!