"Stay" order for signalling organelle

The signaling organelle doesn’t have the most useful function for me: A “Stay” order. This appears in other AI ordering games.

Even if I set the slider for my cells to be mostly Sessile they will still move a lot.

This is a problem because of several reasons. First and most important, they are wasting ATP and Glucose by moving around. This causes them to die, which is a Bad Thing.

Specially egregious, if these cells are photosynthetic cells within a day and night cycle…
(there should be options for the day and night cycle to be shorter, and longer on polar regions; aren’t we speaking of hundreds of millions of years of evolution anyways? Well, that’s kind of off topic)

TL;DR certain cell designs require you to be able to order them to save energy. This would be achieved by telling them to stand still (specially during night). Cells that move ceaselessly and mostly uselessly, will die. Nobody likes suicidal cells (*apoptosis is for multicellular organisms).

The Stay order could also have other uses; for example, not bumping into allies if you just need to pass and there are lots of them. Even for them to stop taking your resources.

1 Like

This is planned to be addressed with AI changes to make the AI take day/night into account.

I’m not sure this is the right way to go about this, because for one this doesn’t help AI species at all so they would still be dying en-masse. And secondly one species member commanding all the others to not move kind of seems like a questionable chemical signaling behaviour that wouldn’t evolve naturally.

Anyways, since we are on the topic of the day/night cycle and you have in on schedule; as the game itself concedes this makes photosynthesis pretty hard. It would be useful to have a sunlight multiplier for the planet including over 100% (the starting blurb says it’s an alien planet anyways, so it isn’t realistic that they all have the exact same earthlike sunlight.

Yes, it would be pretty useful if the behavior to conserve energy at night for photosynthetic cells was hardcoded into the AI.

While I respect the realism argument when it’s relevant, you yourself said many times that the game is based on making concessions for the sake of gameplay. Let’s be real, basically none of the functions of that organelle are realistic, otherwise your cell would be receiving orders as often as giving them. I’m pretty sure the AI doesn’t have any idea how to use it either.

I do think it evolves naturally anyhow, as well. At least in animals, they do signal each other to stay still. For example, smelling or perceiving fear in a herd can trigger this behaviour of alertful stillness. In the same way, unicellular colonies could signal for distress and energy conservation.

On the other hand, it doesn’t really matter that much for AI, this is for player organisms. AI cells, even player species, practically always follow the auto-evol simulator for populations (and AI species come and go randomly anyhow), which means the in-stage suicidal tendencies and how to stop them are mostly relevant for the player only. In fact, most AI cells are terribly constructed and die anyways. Given the evidence, wouldn’t you agree?

The plan for day/night cycle is to properly balance and finish the feature. The reason the warning is there and it isn’t enabled by default is that the feature is not complete:

The AI code is multiple years out of date, because it is very rarely that anyone volunteers to update it. Internally I wrote the feature so that AI cells can emit the same signals as the player just fine (and the closest signal source will be “smelled” by the others), but there’s no AI logic for them to activate that.

Turns out that there wasn’t an open issue for that, I just found this one about the night:

So I opened these 2 now, but it’ll probably take years before anyone is willing to work on them:

This is another thing that hasn’t been done, because there’s way more ideas than people willing to help out with Thrive:

Also the auto-evo is extra kind to the player currently:

I don’t agree with the conclusion that the signaling agent should be a way for the player to prevent their species from moving. Things like the behaviour editing sessility and AI updates related to the day/night cycle should be the primary ways how this problem could be addressed.

1 Like

Thanks for the detailed answer even if I don’t entirely agree with the conclusion on the original issue of the thread.

It is good to know some of these things are waiting to be addressed.

Since you are, I believe, the lead developer, could you clear out something related to this? In regard to AI cells dying from exhaustion.
Even if AI cells often die of malnutrition in the simulation, is there an Osmoregulation multiplier bonus to them or something similar? While many of them are credible predators as far as their function goes, I find it hard to believe that they can last even as much as they do. The big cells tend to have multiple pilum, cillia and flagelli, which would not often be very realistic for a player cell due to the cost/benefit ratio. Then again, it could just be the fact that they pop from nowhere, meaning that it doesn’t matter much if some of them aren’t so well-tuned for long-term survival.

Back in the main, While I agree that the AI should be the primary driver for the cells not dying, I think the signalling agent having that function as well is still realistic enough. Furthermore, it is much easier to code, I believe, and just adds another command to the wheel and a simple directive. It is more important when you move to a new patch and its you and your sister cell, that sister cell and/or their descendants not dying is pretty important (also if you do it changes if you would have a second life or not). Well, just throwing that out there.

What do you mean by “osmoregulation multiplier bonus”? The way auto-evo works (which mostly determines population) is that it calculates the energy sources a species can use and then divides that by the cost per individual to arrive at a population. There are massive penalties in the auto-evo energy source calculations if a species has a negative ATP balance, which prevents completely unviable cells from being generated.

The AI code is one of the areas of the game that I have not actually written.

I can already imagine all the gameplay cheese that players are going to make, after all everyone would really like the option to cannibalize their species…

personally, i feel like the player should be given blocks they connect to make their species’ AI, and the AI of the AI species gets connections randomly made and destroyed and blocks randomly placed and removed. that way, the AI will make itself be able to use any value/function there’s a block for. maybe even give the AI species blocks the player cells don’t normally have access to. the cells are already procedural, so maybe their AIs should be too.

obviously, this would come with the problem of ‘too many species = ram explosion’, but that can be fixed with optimization or limiting the species that can inhabit one patch.

It’s much easier to cannibalize your species if you make them come to you. And this order is already implemented. Standing still is nothing special. Plus, you can’t really damage your own cell species. You can wait for them to die out, but they will die out faster if moving (specially towards you), after which getting their organelles and ingesting them is easier. The cheese is already there.

Good to know, but, I was actually referring mainly to the in-stage cells. There is, for example, an osmoregulation cost multiplier in the difficulty settings, but strangely it only applies to player cells according to the description. This may mean there is a different multiplier for the osmoregulation ATP cost of AI cells (which may or may not be 1).

PS: Here is a screenshot of my cells being a bother to move (and also dying from movement at night). Notice how auto-reproduce is disabled too (in regards to the other topic). Them moving around and blocking me makes it hard for me to find the phosphate clouds.

1 Like

It only applies to the player species:

(Thrive being open source allows anyone to figure out anything for certain about the game by reading the source code)

AI species don’t have their osmoregulation numbers adjusted. It is a difficulty setting to make things easier for the player.

One thing is that microbes don’t actually sleep (unless they are dependant on species that do sleep: https://kids.frontiersin.org/articles/10.3389/frym.2019.00045), because they don’t need to.
If photosynthetic species gain their energy without moving, why would they do that, other than fleeing from predators (assuming their structure would allow it to outrun it, that already might limit photosynthetic capability. If species surely can’t outrun predator than it’s better to switch to defense).
On the other hand, if a species get its energy with mvoing around, there is not much reason not to other than exhaustion or something else.
While staying in place might maybe be something fun, scientific accuracy-fun dilemma kicks in.

1 Like