 That's about as well as it got to work so far. Hi everybody. We are at Artificial Life Creation T Plus 2, the multicellular menagerie. The goals for today was somatic cells, stick around. That's what we just saw not really happening. To get the first living computation foundation book published, talk about that, have more big fun. February was pretty busy. There was a lot of successes, but there was a lot of frustrations. I probably didn't stop and enjoy the fun as much as I should have. I have to keep working on that. Okay, so the schedule. Last time Kill Zombies increased code density. It did pretty well with that. Till this time, somatic cells, body cells, we sort of did that. We obviously need what's coming up next, coordinated movement. Somatic cells. Any cell of a living organism other than the reproductive cells. Okay, now that's pretty obvious when you've got trillions of cells like one of us, but when you're down at the level of single cells or small collections, the distinction between reproductive cells and every other cell is not quite so clear, but exactly that's what we were trying to do with the fins on the fish. In natural biology, there are these cell adhesion molecules. I mean, there's been tons of Nobel Prizes awarded for people figuring out how cells interact with each other, intercellular interactions, recognition, communication, and how they exert forces on each other so that they can stay together unlike the fish that we've got so far. And the model that I'm using for that or at least part of it is this intercellular goo, which I said I would talk about this time, last time, and I'm still not going to talk about it a lot, but essentially what it is is the diamonds that we had before that they sort of had this blue edge. And the new ones that you see in the opening video, they have this kind of brown and yellow stuff around it. That's the intercellular goo. And part of its reason for existing is to create a bigger buffer so that we can be aware of other cells rather than just trying to get so far away we can't see them at all because if we can't see them, we can't communicate with them. So we want to be close enough to have some kind of interaction while still having enough room to get away if things start going wrong. So that's the intercellular goo. That's going to have to develop a whole lot more than it has. Oh yeah. And this was something that I made up ages ago when I was having some tweets with with Michael Levin, you know, people tend to focus on the cells, people tend to focus on the neurons. But really the intercellular interactions is so complex and so powerful that this is a little bit about what we're trying to do here. So, okay. And my stream deck is getting unreliable. I'm going to have to buy another one or do something else. Drivers and driving. So, you know, when we talk about distributed systems control and folks like me who are doing the sort of bottom up stuff are very anti centralized top down, fixed control structures and all that and want to have this distributed system control. But nonetheless, it's still going to be the case that, you know, if you want to get something done as a little collective group, there's going to have to be a single decision. And the classic example in the diamonds is growth and movement. If we're going to go east, we have to all go east. So there needs to be some centralized decision, some, some decision held in common. However, that is established. And you know, in the way it actually comes down to it is that, you know, even though it's distributed system control, there is a point of centralization for any given decision point. The real question is, is that central point fixed by the architecture never changes? Or is it more fluid? Can it be here there or whatever? So in the case of the diamonds that we've got the diamond route, the HG Adam right in the center controls growth and movement, which makes sense because it takes signals in from all the edges about, you know, clearances and contacts information from the goo. And it makes a decision and it broadcasts that out to the whole diamond and they all go one way or they get bigger, whatever it is. But once we have that location where control happens, the controller can become the controlled. And that's what this driver is about. So now we've got this new API. I really want to call them appies. I really want to change the pronunciation of API, but another story. A new new API call advise route that the diamond controller does it looks around itself when it's time to make a decision about movement or growth. And it says, is there a controller nearby to advise me? And if the advice route returns false, then we just go ahead and do whatever we were doing before. So I use this diamond controller to make a simple little driver for a diamond. And they are down here. So this is a comparative one. These things are identical until there's a reproduction. You probably won't be able to see it. There's a tiny little red dot here and over here next to the root in the ones on the bottom. That is the controller that's getting the delegated from the route saying you want to do something. And what this controller says, if I am a right daughter, I want to go east. And if I am a left order, I want to go west. So they spread out. And look at this guy, you know, he the being this being is really trucking. Now of course, once we have a left order of a right order and a right order of a left order, they're still pointing at each other. But nonetheless, these guys managed to distribute themselves a lot quicker than the traditional less motivated just sitting there vibing diamonds that we had before. So that's an example of what's going on. And the fins on the fish are using these drivers to try to stay to in the direction that which is where they thought their their mother was. And now of course, there's, you know, there's all kinds of troubles with that not working out. And once again, I can't get back to my slides. All right, there we go. Big science. All right, much of what I spent the time on for the month of February was collecting data for the a life paper for the a live conference coming in June or something that the paper deadline is March, March 13, the final extended no more extensions is six days from now. A lot of writing to do. But you know, if it's going to be a paper at all, we have to do an experiment, we have to gather some data do some kind of case study. So the run on the grid that we actually saw starting last time in February has been going now for almost 28 days, something like that. Close no less than that three weeks anyway, something. And so I wanted to reduce all that data and answer questions like how long do diamonds typically live? How many kids do they have? And so on. And the data to work with is the same data that we've been watching the camera shots of the entire grid. So I spent a lot of time learning about Python and open CV, open computer vision, a big library, you know, and you see all these demos out in the world full self driving, you know, where there's boxes around everything, you know, person, person, Starbucks, and so forth where the computer has recognized all the stuff in the image. So I figured, you know, how hard could it be to just recognize the diamonds? They don't have that much variation. Well, you know, I'm not a computer vision person. And in the end, I ended up with this thing, which is actually a manual semi automated way of tagging the data. Now this one, it's got all these numbers 102 127, those are all just local IDs that have been made up. So if I create a new diamond, you know, like here, there, you know, 134 or something, and I can move it around. But you know, I'll just get rid of it because there's nothing there. So all these diamonds down here, I labeled by hand. But we can we can watch the whole thing run. So this thing has now already been. So the idea is I label the births and deaths of where everything starts out, and the tracking tries to keep track of the diamonds as they move. And it does an okay job, you know, but if we can rewind it. And so this is what it looks like. Most of what's going on there is is is tracking that the computer vision that I wrote is doing. You can see in various places, it doesn't track very well. Like this thing keeps getting the wrong size. It turns out that down in Lotus four for whatever reason, things are much more finicky. The diamond tracking keeps losing the grid and flying away, and so on. In any event, I've now gotten 20 days worth of data labeled and produced into spreadsheets and CSV files. So that can start answering the questions about lifetimes and fertility and all of that stuff. So that took a ton of time in refreshing my Python so much so that now that I've gone back to I keep forgetting my declarations and semicolons, you know, thanks Python. Okay, so that's that. And finally, companion at caring. This is this is the book I've been talking about on and off for months and months and months written by my dad. And here it is. You know, we've actually got a couple of copies that have been printed and sent to us. This just came Saturday, I think. And here's my dad there. He died in 2008. He was a philosopher and ethical philosopher. And this was kind of his magnum opus, really. And so I thought, I thought I'd just read a little quote from it. So to get you the idea. Now, of course, folks that are following the T2 projects are not necessarily ethical philosophers. But I think we are a fairly special crowd, actually. So let me do this. This is from the end of chapter two, a section called Ethics as Networking. I already like it. Okay. We are morally estranged from our classical tradition. We no longer share its moral ideals, perfectionism, absolute goods, hierarchically organized society, personality conceived as a struggle between reason and feelings. But this does not mean we are immoral. We have a different set of values. It is not so much that we have rejected traditional values as that they have gradually become irrelevant to our lives, our hopes and our problems. The society in which they originated no longer exists. So that's my dad. And, you know, at the moment, I've got copies printed, but it's not in Amazon, Walmart, Barnes & Noble and so forth. So we can't buy it. And it's super expensive. I'm pricing it at 25 bucks for this little thing. It's not about the price per page. It's about wanting to know some more of the background and maybe being interested in that. And mostly, from my point of view, it's just about having my dad's stuff out there. I'm so proud of him. And so that's it. Next month, April 4th, the A-Life paper will have been submitted long before that, like six days from now. But hopefully we're going to get coordinated movement. Let's see if we can get the fins on the fish to stick to the fish and actually be able to buy a copy, not in physical stores necessarily, but online, and have lots more big fun. And that's it. And thanks so much for dropping in. And I hope to see you in a month.