What would a game look like if the most important thing you built was a better starting point for someone else?
That is the question I want to explore with LEGACY. The project brings together architecture, design tools, generative AI and the possibility of more sustainable futures. A game feels like an interesting place for them to meet. It allows us to make decisions, watch consequences unfold and try again. It can also make us care about a place that doesn't actually exist.
The starting idea is a regenerative design game. Players would shape a world through what they choose to build, restore and support. Cities and infrastructure would matter, alongside the landscapes around them and the lives they make possible. Different players should have room to imagine different futures.
These early conversations are about what that experience could feel like. The development work comes later. First, there needs to be a reason to want this world to exist.
A different reason to keep building
There is something satisfying about watching a settlement grow. A few structures become a neighborhood, paths become networks, and earlier decisions start to produce something larger. I would like to keep that pleasure in LEGACY. Building should still feel good.
The question is what we build toward. When progress is mainly about producing more and occupying more land, environmental repair can become another task on the way to expansion. Plant a few trees, balance a number, continue growing. I want regeneration to have a deeper role in the choices themselves.
That could mean asking whether a place becomes more able to support life over time. A project might meet an immediate need while improving conditions around it. A settlement could become more resilient through its relationship with water and landscape. Maintenance and adaptation could be worthwhile activities, with their own rewards.
Players still need freedom to experiment and make bold moves. The interesting choices would be about where to intervene, what to keep, and when a smaller change might do more than a spectacular new object. If improving one part of the world makes another more vulnerable, that relationship should be visible. Otherwise, “regenerative” is just a label.
Playing through a design brief
One direction in the conversation was to give players problems they could approach as designers. A place needs something. There are existing conditions, limited resources and people who will live with the result. The player works through possible responses instead of simply selecting the next unlocked building.
We discussed floating settlements, restoration projects and a floating hospital. These invite more than one answer. A hospital could be considered through how people reach it, how it connects to its surroundings and what happens when conditions change. Its appearance would sit alongside those practical relationships.
I can imagine a mission starting with a short brief and a place to explore. The player looks at what is there, puts together a proposal, sees how it works and adjusts it. It has a familiar design rhythm: understand, propose, test, revise.
It shouldn't require an architectural education to enjoy. The challenge is to make the feedback readable through play. A route becomes crowded. A public space attracts activity. A restoration project gradually changes its surroundings. The player notices something and forms an idea about what to do next.
The early vision moves between the planet, cities, individual projects and construction mechs. That scope will need narrowing when we test it. But the connection matters: local decisions belong to larger systems, and zooming out should reveal something about the choices made close up.
AI as someone to design with
AI generation could help players turn an intention into something they can respond to. They might describe a floating neighborhood with shaded walkways, or a restoration mech inspired by marine life. An AI studio could offer concept directions and compatible parts, which the player then selects, combines and refines.
There should still be something interesting to do after the first image appears. If a prompt produces a finished solution, there may be very little game left. Generation could instead offer a starting point, an unexpected possibility or a way to express a preference that the standard menu doesn't quite capture.
The mech discussion made this more practical. A completely new body for every prompt would create problems for animation and consistency. One proposal was to keep a defined skeleton and compatible components, while allowing generated textures, colors and decorative details. The machine could feel personal while remaining something the game understands.
Architecture could follow a similar approach. Curated components would provide the structure, while generation helps explore their arrangement and character. It offers freedom within a system whose behavior can still be explained and tested.
The game rules would need to remain responsible for outcomes. A generated design shouldn't perform better simply because its description calls it resilient. Any benefits must come from defined mechanics, so players can understand their decisions and compare alternatives fairly.
The core experience also needs to be enjoyable without generation. AI could add individuality and surprise, but building, restoration and decision-making should already be worth doing. That seems like a useful test of whether it belongs.
A future close enough to care about
A regenerative game needs a sense of time. Some decisions help immediately; others take longer to show their value. Some create obligations that only become clear later. That delay could give choices their weight.
The urgency I want to explore comes from those relationships. Waiting might make a problem harder to address. Committing everything to an immediate solution might leave fewer options for the future. There should be a reason to act and a reason to think before acting.
I wouldn't want a constant alarm. A world made entirely of emergencies leaves little room to notice what is worth protecting. The visual discussions leaned toward a living illustration: atmospheric places, readable compositions and small signs of activity. Time spent enjoying a place could make its future matter more.
Construction mechs could give the work some character, too. A machine the player has shaped and used across several projects might become associated with the places it helped change. Familiarity could be as rewarding as customization.
Another proposal was a gallery showing the world at different years or milestones. Looking back, you could recognize a decision: that was when we changed the waterfront; that project connected those neighborhoods. If AI helps create these images or stories, they should reflect what actually happened during play. The interesting history includes compromises and things that didn't go as expected.
Why the name LEGACY matters
LEGACY is about leaving behind a regenerative and better world for the next generation. That is the simplest explanation of the name, and the idea I would return to whenever the project becomes too complicated.
The word asks us to look beyond completion. A building can be finished and a mission can be won, but the world continues afterward. Someone inherits what we built, along with what we used up, neglected or made possible.
I would like that inheritance to be more tangible than a final score. One possible approach is for later stages to begin with conditions created earlier. Players would encounter their own decisions from the receiving end. Something that once seemed convenient might now require care. Something that took patience might have opened possibilities that weren't available before.
That remains a proposed direction, rather than a settled generation mechanic. But it gets close to what the name means to me. Success would include the choices left to those who come next: useful resources, healthier places, functioning systems and room to find their own way.
A better inheritance wouldn't mean a world where every decision has already been made. It could mean one that can adapt and recover when circumstances change. There is a worthwhile question there about when to intervene and when to leave space.
The name also brings attention back to what lasts. Players may create memorable cities and machines, but the longer story belongs to the world they help sustain. A meaningful result might be something that keeps working after they have moved on.
Keeping the idea open
A meaningful theme doesn't automatically make a compelling game. We still need to find out whether the choices are interesting, the consequences understandable, and looking after a place enjoyable enough to return to.
There are questions about the model behind it, too. How do we represent ecological relationships without pretending to simulate everything? Can different approaches succeed without prescribing a single ideal city? How do we give players useful feedback without reducing every kind of value to one number?
Those questions are part of why LEGACY feels worth exploring. A game lets us make a possible future visible, act within it and see what our choices ask of us.
For now, I want to keep the starting question close: what kind of world do we leave behind? If LEGACY can make that enjoyable to explore, and personal enough to stay with someone after they stop playing, it will have a reason to exist.