 The T2 Tile project is building an indefinitely scalable computational stack. Follow our progress here on T Tuesday Updates. So the Artificial Life Conference was last week. I've been aiming at that for over a year. We were originally hoping to have a whole grid all set up and running. It was supposed to be in Montreal, became virtual online because of the pandemic. So I had my first Zoom experience. But it was really quite impressive all along. Monday morning, bright and early, the very first event of the conference was Functional Programming for Artificial Life, a tutorial by my friend and colleague Lance Williams. He is a big functional programming and Haskell fan. So the tutorial was all monads all the time. But by the end of the time, he was showing us this absolutely crazy stuff, these roving piles of chemical-like things that interact with each other according to the rules of functional programming. Combinators like lambda expressions and so forth on super steroids. Thing has a mother with a program that copies itself, extrudes it into a daughter and then pinches it off, does the whole thing. And that then continues. It's absolutely crazy stuff. And it's implemented in stuff that's pretty close, not quite, but pretty close to stuff that you could imagine doing in something like the Movable Feast Machine. I did my tutorial on Monday as well, programming software with Splat and ULAM and ended up being mostly ULAM because I was working bottom up. I was characterizing the Movable Feast Machine as a chaotic good cellular automata. Trying to drive home the idea that it's not merely asynchronous, but it's viciously asynchronous. Things can come and go. Things can get blocked and then start up again and so forth. Did examples like Hello World, which was pointing out that it's really like Goodbye World, right? Because it's a typical Hello World program. It does one thing and then it exits. Living computations don't exit lightly. Did 2 plus 2. Why do 2 plus 2? Because you say int x, y, z, well, you can't do that and ULAM it takes up too much space. 32, 32, 34, that's 96, haven't got that many bits. Eventually worked our way up to the swap line, which was going to tie into the lightning talk that I gave at the end of last week, which I showed first in ULAM and then also in Splat. And Liza Shulyayeva, I guess was at the tutorial and put up, she called them hasty notes, but they're really quite impressive. They go on and on and really get pretty much the main points I was trying to make. The goal, useful scales, indefinite scalability has a personal note about don't we already have indefinite scalability? And certainly I was going too fast and trying to cover strict indefinite scalability takes a little bit of extra unpacking, which I didn't do. So when you say just restart your program, you have two cases. The program sits on one host and does its job. And if that's the case, then you're limited by however many address bits, the fit and RAM of the machine that you're dealing with. And if it's a network program that's designed to scale out on the IP network, well, then you're limited by IP address space. And people think, oh, well, IP abstract is really big. It is really big, but strict indefinite scalability is actually more of a design philosophy that just when you actually cash it out in specific finitely scalable cases, you always run up against something like 64-bit address space of RAM or 128-bits of address space for IP6. And the limit itself is less important than the fact that if you try to get away from it, it pushes your whole design philosophy in another direction. But these notes are great. Oh, and we grew a new section in the Living Computation Foundation.org, Living Computation.org website for education and outreach. Compared to other living machinery on Earth, the human species has had the success it has had primarily because of its abilities and propensities to educate and learn from each other. That's really the bottom line. And that's why the Living Computation Foundation has this two-prong mission. Research and development, like building the T2 tiles, and education and outreach to spread the information about there being another way to compute that we need to start building up engineering and science body of knowledge about how to compute the Living Computation style. So in the education section, we have a thing for the tutorial, memorializing the tutorial, the code samples and scripts and so forth for anybody to try it on a virtual machine, a Ubuntu virtual machine or a real one, if one has it. The slides are up in PDF and the YouTube video of recording from the Zoom is linked there as well. Yes, and finally, working on the website, I got around to finally running the thank you script. So our latest Living Computation Foundation nerds, Damian and Jason, thank you, should hopefully have their official nerd numbers sent out to them. We're into the 230s now. Another tutorial was the Bibbets by Leo Casson. I didn't even know how to, I'm not sure I'm saying his name right, I didn't even know how to say the Bibbets until I heard his tutorial. And this was good because, you know, so as he says, he's not a researcher when he first started the project, he didn't even know that A-Life was a thing with a name. He was just doing it because it was cool. And a lot of people come to A-Life that way and it's really good. It's great. And one of the things that's really important about the artificial life conferences is trying to make room for as many of those new folks to come in as possible. And Leo doing a tutorial is exactly an example of that. You know, it took a bit of guts on his part to come on out and do it. But it was great. And so he's doing a thing where, you know, you directly engineer the genes that go in your creatures and, you know, sort of in the spirit of Steve Grant's creatures from the 90s and lots of other systems that people work on. It's the typical way that people, when they first come at it, work on it. And it's great. And really, you know, even if there's evolution within the creatures that are being simulated, the larger scale of evolution is the developer, as he says, continuing to add wow factor and continuing to edit the thing and develop new stuff. And, you know, he's doing five-digit subscribers, six-digit views for his videos. He's doing really good. So that's good. That's education and outreach of a whole other sort. He kind of, he stressed the idea of having it being an ongoing thing versus the sort of one and done, one paper, one thesis, one so forth, which, you know, I really agree, is a problem with artificial life. You get these, you know, single starts that explore one idea and then poof, they're gone. And, you know, he calls out Splat and the T2 tile project as examples of things that are ongoing. Thank you. You know, one view of it is it's a long-going project because it's a long-going project. And one is it's a long-going project because I make progress so slow. But I totally agree. The benefits of a community, especially for accountability and motivation, you guys and you beings, you folks, you really helped me by letting me know you're there making comments and so forth. So I super appreciate that. And hopefully, you know, we'll continue to have progress for a while here. And out of that, so I've talked about Mike Eleven before. He did the first keynote of the A-Life conference and it's just more crazy stuff. Why don't robots get cancer? I mean, he has this thing where, you know, you have biological models where they have cancer tumors forming and by screwing around with the bioelectric fields, I mean, it sounds like the matrix, but changing the ion channels to affect the thing, the cancer goes away. I mean, it almost sounds like New Age Wu, but Mike Eleven is hardcore. It's great stuff. You know, and he's lots of crazy examples. You cut the tail off a salamander. You graft it on where the leg's supposed to be. The tail turns into a leg all the way down to fingers. And he shows some of these two-headed plenarian worms that they make by cutting them apart and stitching together and messing with the bioelectric ion channels so that the developmental process gets confused about which end is the head and starts growing heads at both ends. And, you know, in his talk, this is video. So you can see the little goofy little googly eyes at each end of this thing. It's really kind of weird. But in the video, this thing is swimming around. And, you know, how does a two-headed creature swim around? Swims around in a U-shape, mostly, although it seems like the bigger head tends to dominate when they actually end up going one way versus the other. Similarly, you can take embryonic eye cells and graft them onto a tadpole's butt. And it will grow into an eye that works. How? Because the eye cells put out an optic nerve. The optic nerve cells look for the brain that they're supposed to hook onto. Don't find it. Find the spinal cord. Make active synapses onto the spinal cord. And the whole nervous system eventually adapts to use those signals so that the eye is providing information about lighting behind you. This is exactly the kind of thing that we're exploring by trying to say, how can we retake a high-level goal and re-engineer it into something so that you can make progress locally? You can wake up. You can look around at yourself and you can say, well, based on what I see here, I should grow in that direction. That's exactly how embryology manifestly clearly works. And it's only because we have this kind of conscious thought model where we think everything stays in one position, everything has to be told what to do, that this even surprises us. Really, we should be expecting this. And the more I get immersed in bottom-up, multi-scale, agency at all levels, like agency at the level of embryonic eye cells, making decisions based on local conditions about the smartest thing. How can they most help? Well, clearly we should grow and hook up to the thing that seems most brain-like. That this is just a consequence of making a complex high-level, high-order goal, breaking it down into low-order accessible components. It's great examples all through it. There he is himself. He was in a second workshop later as well, more stuff that you don't even have to get the number of cells right. And even if you have only one cell, it'll make a tube around a whole thing. Great stuff. There's Josh Bongard. This was in the later workshop. They're trying to set up a Proteus Institute, which is aiming at trying to be able to transfer learning between radically different systems. It's very ambitious. I pushed back a little bit on it. I mean, I think in principle this can be made to work, but trying to do this on top of traditional deterministic computing is gonna be a big lift. But they're hoping to be able to transfer from like soft simulated voxel robots to Melanie Moses swarm robots, and get a learning speed up by crossing species in effect. It's great stuff. I did my lightning talk. This is five minutes. It's up on the Dave Ackley channel. I encourage folks to take a look at it. I sort of include it in our T Tuesday update by reference. And it went okay. I did it using OpenOffice instead of this software that I usually do. And OpenOffice has a better ability to include little movie clips. And that's one of the disadvantages of this visualization software is it doesn't. I think I wanna work on that. But the lightning talk went okay. I built up to showing stuff that wasn't even in the paper where we could do have a little local stuff that was sorting itself out locally and then having passing computations across it and then having the entire computation moving down the grid as that was happening. And it got it working. But that wasn't even it. There was a whole bunch of art exhibits. This was the People's Choice Award submission. I didn't vote in the People's Choice Award because I had gotten roped in to be in one of the official judges. There was lots of really interesting stuff. And every dimension from sort of ambient sound, recordings in natural environments to explicit sort of a-life electronica to quilts that had living system, mechanical life interconnections and so forth. The Art and the Science Award went to this thing called Blind Painter that we liked by Mitsuyoshi Yamazaki. There was a ton of great stuff. There was a Minecraft server specifically for the conference. And you can see the topiary robot horse from the conference poster now exists in 3D in the Minecraft server. Somebody carved Ulan into the side of a mountain in the Minecraft server after the first day of the tutorials. Thank you. And really everybody, the team that organized this, Josh and Juniper and Lauren, they were working nonstop whole week, especially Juniper. She was everywhere. She was the conference mom, not just doing technical stuff, but also asking questions and hurting people from one lecture hall to another and so forth. It was remarkably well done. If anybody could keep up with the sort of thing that Juniper did, it's probably Jitka who's running the conference next year, A-Life 2021, which is gonna be a combination of physical in Prague, in Czechoslovakia, plus virtual online. We certainly wanna have the Living Computation Foundation, the T2 Tile Project having some significant visibility there whether we actually go to Prague or not depends on the state of the world. But who knows, we shall see. So that was A-Life 2020. Going forward, I wanna get back to the work of Native Engine. We've still got bugs. I wanna build a second power zone because that's the last untried thing, is having two separately powered grids, sub grids that are connected together that should be exchanging data even though they're not exchanging power. And if I can dare myself to try to get down and figure out what's going on with those guys that won't reboot if they're in the middle of a powered up grid, but they'll all power up okay by themselves. That's hopefully make progress on as many of those as possible by next time. And that's it. The next update will be August 4th. I hope you guys or you folks are doing okay. Thank you so much for being here, helping support the project. Hope to see you next time.