What is it?

A world/society simulator with nested hexes, down to the individual person/tree and out to a whole continent. multiple people can form factions (or leave them), trade, go to war, claim territory. forage, chop trees, grow food, research tech trees. hierarchies are emergent based on individual and group choices, "borders" aren't real. buy loyalties are. I'm currently having my mind blown. development is moving stupidly fast. we'll see where it hits the wall of "idea sounds incredible on paper but isn't feasible in software or UI" - haven't gotten there yet!

a mix between travian and minecraft, the game is making the game seems 👍

yea. or like dwarf fortress + civilization with a huge helping of anarchism 😆

it's come a long way in a day....

This is way more ambitious than my ant game! When can we play?

i'm "playing" it every hour of the day as i'm testing

it's regularly blowing my fucking mind, btw.

And it's not just AI psychosis? You don't seem prone to that, so my interest is piqued.

I'll wait to see more before asking more questions.

so far, no. i'm playing the actual game and it's actually fun and shows true emergence/complexity. the game itself uses no AI whatsoever. it's all just Rust code for complex simulations. so the "ai psychosis" risk would end at my talking to claude to generate the code. The static, AI-less code is what i'm interacting with when playtesting, and that's truly cool. The one place where AI may be hypnotizing me is with promises of this game model being able to scale past early agrarian society. I've (and I mean **I**. not claude's idea) came up with an abstraction/folding system where the same rules that apply to simulating an individual and then apply to a household - which is a folding of many individuals together. and multiple househoulds form a band, many bands/neighborhoods can form cities, cities -> nations, etc. and all throughout, the "folded" entities still retain some emergent surprises - like there is a probability of a single individual "poking out" of a city if the probabilities align so that it makes sense for him to do so. My rule with the game was that any level of organization, down to the person, should be able to emerge and meaningfully dissent/break away, if they have a good reason to do so and the means. So far this has served well at the levels of scale i've got: individuals get fed up with their band's location far from firewood, so they might choose to take their family and start a new band. or they may convince the rest of the band to move at once. and bands themselves sometimes merge when it is advantageous for them to do so. Still yet to see if this will keep scaling up through the ages! But I have good reason to expect this approach to be arbitrarily scalable without losing too much of the detail that makes the lower levels dramatic and fun. I'll post some recordings of feature demos today. i get claude to do these for me sometimes.

Currently developing a system whereby disparate bands meet seasonally at gathering grounds to share knowledge, settle disputes, propose marriages (like the Icelandic Althing). Sites known to people get used over and over, with structures being maintained or falling into permanent ruin for later societies to find. This, too, is a generic concept - "sites" are interesting places that people can discover and can share knowledge of with each other. for now it's these gathering places, but as the ages (and game dev) progress, a "site" might take all different sorts of forms and be shared for any number of reasons. I would imagine that later on, gold prospectors will use the same generic mechanism to share or sell valuable knowledge of gold veins. Fully emergent gold rushes should be a thing or else I've failed.

To wit: So, is this AI psychosis? well, the Sites thing works for these hunter/gatherer seasonal sites, so not hallucinating yet. but will this same generic system work for 1800s gold rushes and Indiana Jones-style relic hunts? Yet to be seen!!

re: "gold rushes":

Bookmarking reply