Friday, December 8, 2023

A New Base Genome: Fear is the Mind Killer

As has so often been the case with me, what I said my next post was going to be wasn't actually what it is. At least in this case it's still on the subject of working on my new base genome as opposed to something completely unrelated.

This time around, it's about a feature I'm porting over from the TWB/TCB genome. Specifically, the pair of emitters that substitute for the original "Grendel frightening" neuroemitter.

First, a brief explanation about neuroemitters in vanilla creatures. What they're supposed to do is emit chemicals based on the state of a creature's brain. In the default norns, it's set to emit fear, crowded, and adrenaline whenever there's a grendel around.

The problem is that neuroemitters just plain don't work. It's something I suspected even early on into my gengineering career before it was confirmed more recently. It's a bit of a shame since they had the potential to be extremely powerful gengineering tools that are easier to grok than editing the brain outright, but it is what it is.

Fortunately, the default neuroemitter is fairly easy to replicate with normal emitters, as was the case with the TWBs/TCBs. Let's take a look at the genes in question.

For the uninitiated, emitter genes emit chemicals based on the physical state of the creature. This isn't limited to the state of the brain; it can also be things like whether or not the creature is underwater. It's just that in this case it's set to fire based on state 0 of neuron 37 in the Stimuli lobe.

If that doesn't make any sense to you...well, it didn't to me either. Fortunately for me I was able to locate a thread on Creatures Caves that explained things well enough for the purposes of understanding what's going on here. The short version is that the Stimuli lobe is where the combined input of the creature's senses is stored (in order of importance: what the object is, whether or not the creature can see it and/or smell it, and how fast it's moving if it's moving at all). Neuron 37 refers to grendels. So in short, it's determining if there's a grendel in the proximity.

No word on what state(0) means, but given that there's more options for neurons than there are neurons in the stimuli lobe and that they all have options for states(0 to 3), I think that always looks the same regardless of what tissue (lobe) you pick.

When I was sure I understood what these two genes were doing, I copied them over to my base genome and did some testing. At first I left the values as they were; I don't think the TWBs/TCBs changed up that much that they wouldn't work for 2017 creatures as well.

The results...well. What I was hoping for was something that'd just make norns wary of grendels. Enough to give them a headstart in fleeing when staring down an aggressive grendel, but not enough to compromise their normal behavior if the grendel was friendly. What I got was norns freaking out even when the grendel was minding its own business.

As much fun as it was to watch (it's especially interesting seeing the fight-or-flight genes getting a kickstart; though I'm working with just the ChiChi genome, seeing the otherwise nonaggressive ChiChi Norns become grendel slayers when they're backed by a group was definitely interesting), it was a little overtuned for my liking. It should be noted that in 2017 creatures fear doesn't overwhelm hunger (for the sake of crossbreeding), but it was getting high enough that the fear gait kicked in and made them walk away from things they wanted to interact with. 

I considered playing around with the values in the emitter genes, but...while I may have a better understanding of what the locus means, I still don't entirely understand what effects it has on the values in the emitter itself. Particularly the threshold, a.k.a. the value the locus needs to be higher than in order to do its thing.

I played around with those values anyway, before deciding I'd be better off with a different solution that I understood in full. And this was my solution:
Chemical reaction genes are pretty self-explanatory, I think. They're quite possibly the easiest genes to understand, save for maybe the genus, pigment, and appearance genes (and that last one can be confusing if you don't understand how the game handles creature sprites), and it's not too hard to get something satisfactory out of them.

This particular reaction takes grendel smell (CA smells are what creatures use to navigate, among other things; Aiko has an entire post over on her blog if you want to know more) and turns it into fear and adrenaline. Effectively, it accomplishes the same thing as those two emitters would, but in a more comprehensible form. The exact values took some messing around with, but ultimately this gives me the wariness but not panic I was looking for. Plus, this gives me something to work with for my base grendel genome; grendels have a very similar gene that makes them angry when they smell norns.

(If you're wondering what adrenaline does in a creature, it has three purposes: it's emitted at low energy to make the creature produce more energy, it'll break down glycogen back into glucose if it gets too high, and it's also used by the fight or flight system to increase fear and anger. The latter is the purpose it's meant to serve here.)

Alas, there is a downside to this. CA smells travel between levels and even between metarooms, so norns with this reaction gene will likely start getting scared even before seeing any grendels. As far as I could tell, the emitters based on the Stimuli lobe only start triggering when a norn is in proximity to a grendel; should the grendel change levels or rooms they stop firing. If you understand what's going on better than I do, that's probably the superior option.

I also considered not implementing a system like this at all; I'm not sure how big of an advantage it provides against aggressive grendels, and it's a downside when dealing with friendly grendels. And also I don't often play with more than one species at a time anyway. But it does provide an interesting behavior for those that often make multi-species worlds, and I do want creatures that are fun to watch. So I figured it was worth having.

Well, I sure had a lot more to say on this than I thought I was going to. But this should be the last of the detours and NEXT TIME will be the full changelist from default 2017 creatures. Until then, folks.

1 comment:

  1. I'm so hyped to see what you'll do next! OMG a new base genome.