Today we submitted to onward essays.
Well, a long time ago I was searching for unintentional asmr on the internet and I saw that someone recommended Dave Ackley’s video about artificial life on reddit.
They were right! It was very relaxing to listen to, but it caught my attention for reasons more than that. My head started to think through what he was saying, about life, about simulation, and how it’s all a form of computation.
So I binged all the rest of Dave’s work.
His largest project, the T2 Tile, didn’t exist back then, so it was beyond efficiency, and machine learning, and his huegene and his demon horde sort. And many more.
And of course — spatial programming. And this one really stuck with me. I couldn’t stop thinking about how you could ‘simulate’ all sorts of things with it. I imagined a falling sand game that’s built on primitives of spatial programming.
And the really exciting thing — at least for me — is that this approach can be indefinitely scalable. It’s not like our traditional computers that encounter scaling issues when you make them bigger and bigger. They become fragile — a wobbly tower of cards that falls down too easily.
No. With spatial programming, you don’t encounter those same issues of size.
Sure, there are always other issues, like finding enough space, or power, or money. But the computational issues are gone. You can trade them away for a completely different style of computing. It’s non-deterministic, built on spatial relationships, with event windows, and yada yada yada.
I wanted to test out if spatial programming could actually work with a falling sand game, so I made a proof-of-concept called sandboys (weirdly gendered). And it worked!
Then I sat and thought about it for a year.
Then I turned it 3D. It worked again! And so the sandpond saga was born.
Andrew Walpole noticed my videos, and shared them in Dave’s community chat. I was lurking there at the time! It had gone full circle, and Dave got to hear someone repeating his ideas back at him.
Time went on and I built out spacetode, my spatial programming language, strongly inspired by Dave’s splat language.
And then I built a new thing, a visual programming language thing called cellpond, that was only 2D, but provided much more expressive power. It allowed you to do spatial programming without escape. And while it didn’t deliver on the full promise of spatial programming’s robustness, it did provide some novel ways to express complex behaviours.
Cellpond has been my most ‘academically’ successful project so far, which is still not very successful in the grand scheme of things. It doesn’t have a paper, or even an essay, attached to it. To me, that doesn’t matter, but I suppose… I suppose that it would help to be able to wave a flag around at the end and say “I wrote a paper! I wrote a paper!”
It would be another stamp of legitimacy under my name, so that maybe maybe maybe —
Maybe one day people will take me seriously.
In my journeys around the various nerd communities of the world, two different people told me to start a wiki/blog in one week. So here we are. That’s why this place exists.
The wikiblog you’re reading got more attention than I expected, and all sorts of people started contacting and emailing me. It was very lovely!
So I encouraged people to do it more, and it completely backfired in a nice way because now I receive more emails and messages than I can ever possibly hope of answering. Sorry if I haven’t responded — I do read and appreciate them all. (The best way to actually reach me is to submit a full submission to torn leaf).
Anyway, before my email inbox explodes, Dave reached out to me asking if I’d like to do a call!
Yes! Of course! How have we not done that already?!
And we called and we chatted and it was great. On a monthly basis, I spoke with Dave and —
As a retired professor with many publications to his name, Dave was able to give me lots of guidance and advice on navigating the academic (and sometimes academish) world. I learned a lot, and I felt immensely thankful for it all.
Then Dave suggested something crazy.
LU: Write a paper!?
DAVE: Yes, a Wilson & Ackley (2024) paper.
And so we thought out some paper ideas — ideas for topics — whatever was in the middle part of both our work’s venn diagram. And we came up with a bunch.
Time passed, and no work happened.
Not surprising, I guess, seeing as how many things I’m trying to juggle. But we needed a solution — some way of making this go quicker.
So we started recording all of our calls, and pumping out the transcript, and seeing if we could turn that into —
Well, I suppose I shouldn’t give out any spoilers, just in case we do get accepted.
We collected together everything in a private repository called “LD10”. (I suggested “Wackley” but “Lu Dave Zero” seemed much more sensible).
We started building out a .tex file, and I learned just how awful that format is, and that it’s pronounced lay teck not latex.
And no! I don’t want to
use a microsoft word template, orsome other closed off systembecause — by writing directly into a.texfile, we were able to see diffs, and do version control and all sorts of stuff in a way that I’m familiar with.
Writing writing writing, editing editing, and — when we needed more content for certain parts, or we needed to address a point, or a criticism — we recorded more and more talks — and I put them into the essay in various places.
We chose an unusual format in the end because it demonstrated the non-deterministic and robust principles that we were both calling for. We were putting our money where our mouth was in a way.
Time went on, and work increased. We really needed to bring this thing into a landing, as Dave often said.
And so the work became less about writing, and more about getting the figures in place, the formatting right, and eeking out every last punch we could find, cutting away the crap, and leaving the core, oh and don’t forget the references!
LU: Are you sure we can use tldraw as a reference?
And today we pushed the button!
Today we submitted this thing to onward essays!
I have no idea if we’ll get rejected or not. I think it is possible, just because of how weird the essay is.
I hope we get accepted. We put in so much work and effort and care, and I think it will find an audience if it does get through. But maybe it’s just too unusual a format or whatever. 🤷♀️
Whatever happens, I don’t regret this experience. I’ve learned so so so much from it. I’ve learned all the little tricks and best practices, on picking your title, crafting your abstract, even choosing the right keywords. On thinking through in detail, and re-thinking and re-thinking, and phrasing and re-phrasing.
When we hear back, I hope to share more about how the writing process went.
Jonathan Edwards told me (and all of us) to write a paper. And, hey, I didn’t write a paper, but I did write an essay, with a great friend and mentor of mine. It was a huge amount of fun, a huge amount of stress — of lost evenings and weekends — and yet also…
A huge amount of joy gained. All those moments of cross-continental cross-generational collaboration are not ones that I ever plan to forget. Not to mention all the things I learned and the skills I practised for the first time.
Would I recommend this for everyone? Certainly not. But I would encourage you to dabble, at the very least. I truly believe that the most satisfying and rounded work comes from blending fields, science, art, academia, fun, pain, and all the rest.
Whatever code you speak in, there’s a Bloody Lot of good that can come from learning a new language every now and then.
and that’s it done, step by step.
I would like to thank my partner Flora for all her patience and encouragement over the last few months. And of course, Dave for all the mentorship and big fun. I also send thanks to all of my paying supporters for funding this research!
Back to the wikiblogarden.