 Hey folks, it's been a month since we created life. Let's take a look at how it's coming along So that's it. It did not look so good for the living creatures there. So My first goal for the last month was the first steps of a better life. How are we gonna beat the zombies for one thing? So You know, it becomes a little philosophical, what is a zombie because you know, basically If it hasn't got the code if it hasn't got the ancestor code running in it We want to get rid of it But even if it does have the ancestor code in it if it's messed up somehow so that it can't actually do something We want to get rid of it also. So I ended up making up these Productivity measures which feels very top-down and not in the spirit of bottom-up and emergent and all that stuff But hey You want to get things done. You don't want a lot of zombies hanging around so at this point there are now two Watchdog timers inside every diamond grid one that's Reset we feed the watchdog when and we actually try to execute an instruction So that means at least we have a diamond sequencer and the diamond sequencer still has some code in it Then there's a much slower watchdog that we feed much more rarely that says have we actually moved and if the Either of those go off we kill the thing we declare it's a zombie We'll look at another example of that a little bit below Also for a better life one of the big problems was that we had The code kept getting chopped by the cell division which has been completely redone but in particular now there's a Program ball lockdown feature that is part of the ancestor code. We can say is the loop whatever loop we're dealing with Is it free to reorganize itself and grow and shrink and all that sort of thing or is it locked down and With the values of how many hops we have away from the diamond sequencer we can get a sense of how How it's spread out the loop is and when it seems like it's tight enough we can lock it down. So here's a little example So this is the way it used to be this thing is flopping all over the place It's a little bit hard to tell because the color changing so there It doesn't die immediately it doesn't die until a thing grows But the corner was clipped and that's why it became a zombie So here the same thing happens in the beginning it can flop all over the place because it needs to grow But now it's been locked down. So this shape what a little raggedy thing whatever it is can no longer change things can move along it, but the shape doesn't change and we have much better chance of getting through Cell division successfully. So that's been a big help This is by the same token the fact that we have to Compress the ancestor code down as small as possible. So it's not near the cell boundaries during cell division Was like, you know, I Look I looked into DNA, you know this about DNA like a human cell is like a thousandth of a millimeter But if you took the DNA in a single cell and stretched it out, it's like two meters long It's completely ridiculous. You know billions and billions of bases and you know The best that I could possibly hope to do is maybe thousands of bits in in this model because it's a very shallow model We each atom is 96, you know, this has got 26 atoms in this loop. This has got 24 But I did a giant refactoring so that now we can have two up to two Codons in each pocket. So if you follow through here, there's one codon in each pocket all the way around This one sometimes it's one. There's two There's two there's two there's two and so forth. So potentially We can double the density although that doesn't work out that well because you need to have slack in the system This has helped a ton put it together with lockdown and we've got much more reliable living The first steps of a better life. So in particular Now what happens? How do we do with the zombies now? So Ancestor 14 234 version 4 sub version 34 Is running at the moment with the double density code the lockdown the auto zombie kill watch dogs and a couple more things I don't even have time to talk about I'll talk about intercellular goon next time. Um, so let's just take a quick little look At some of that going Check it out Those are barcodes zombie auto kill. Yes, nice Those are barcodes representing the parents lineage that led up to the moment of either death or reproduction so major progress And okay second point that Even with the double density and i'm also been developing new code strategies to uh Reduce the number of bits per instruction to make it possible to put more Instructions in a given atom and you know, it's gone from the first one could have four instructions in an atom This is the second version that we had last time could have seven Now we're up to ones that we can have 11 instructions in them And it's a it's a straight trade-off. We're just doing let's bits per instruction So we have to pick which instructions we think are most important to put into this particular coding But put all that together. We're still going to need more space. So here we go I want to try to build lotus four Expanding the ballroom floor After the dancing has already started So I cheated a little bit by putting in the hanger straps here Before the run started. So even if this succeeds, which it won't it won't it can't there's just so much to go wrong All right, we have two fans it's time to put put in the hooks what they look like There's one bunch with a beautiful look at that Or any data only All right, one of them was pretty easy There's two parts. I'm worried about this. This is one other part is The big software update So that is The frontier Standard intertidal connectors and it's time for this Check it out 30 more times and in principle we could do lotus five If we lose no more of these okay so Power supply. Well I think I'm going to use this giant one I'm going to use 780 once here we go number one It appears rebooting number two it appears The dancing has not been disrupted go Look at that Yes That's great, isn't that great? No, uh, it'll be first light lotus four uh Nice bright screens that haven't worn out And this is amazing. This is better than I was ready to hope for up on That screen is not working on the center tile of the lotus always something Actually, it's power this guy first There we go. All right. So basically that's it So there it is. Here's my title card for today Uh my effort to Build out the ballroom floor my second attempt at building out the ballroom for a while the dancing has already started has failed again Uh, there's progress in the sense the first time it failed because of hardware mechanical problems this time I got past it it failed because of software updating and lack of robustness in the software is not all that surprising But it actually did reveal A specific design problem, which I don't think I've been completely aware of that even though the mfm t2 Engines on the tiles they check they use little crypto hash signatures and stuff To make sure that the tiles that they're talking to are running the same physics the way it was designed They don't actually ensure that they're running the same version of mfm t2 And you know, I may have decided I may have thought that was a good idea at the time But now I think it's a bad idea and in particular what it meant could happen Was that it could start up and if it managed to get the new physics for For a 14234 Then it would just start going ahead and just start talking to the other tiles even though it was running a completely different and like two years older version of mfm And it got into all kinds of troubles. And so I'm not at all surprised that We ended up wiping out the dancers again But that's it. We now have 76 tiles On the board. That's four lotuses four times 19 I'm going to live with that for a while. I'm getting worried about power Capability in in the room and we'll see Finally a multicellular menagerie that is What I've decided to focus on for the next six months. It's a half year goal And the idea is, you know, originally I was thinking, uh, you know, let's let's, you know, make make cells that Make a multicellular organism and they they drive a car or I don't know what Something that actually got off of the grid out into the real world And I still want to do that, but it's going to take time to get there. We have to earn our way there That is the bottom up Spirit that is the bottom up fact of life that you cannot design these things just in your head And expect to get all the interactions right. I don't care how smart you are You know, you can do theory But theory is easy compared to all of these inherent trade-offs where there is no perfect solution And you have to say what how can I sort of slalom between these Good bad and otherwise decision points So my goal is in a month by month each month I want to have a new kind of creature a new little multicellular creature to add to the multicellular Menagerie that demonstrates some particular features. So next time in 3137 somatic cells So right now all the cells that we see are germline cells. They all contain the ancestor code. They're all capable of dividing It's time to figure out ways to peel off a little bit of code and say, okay, this is your controller You you're just a you know a skin cell a hand cell a body cell a tail. Who knows? Um, and you get a little snippet of the code, uh, which, you know, normally you Doesn't matter how you do. Oh, we're out of time. Oh, no So that's the plan, uh, um, also the first book published and have lots of fun And uh, thanks so much for stopping by. I hope to see you in a month