Section 01The Core Thesis
What people want and have never gotten: a world they can live in, that feels both real and nuanced up close, and vast and persistent far away.
The gap sits between two design philosophies that have historically been treated as mutually exclusive:
Rockstar's Approach
Beautiful, cinematic, emotionally resonant worlds with aggressive LOD streaming and bespoke animations.
Click to see the trade-offThe Trade-Off
Heavily authored, hand-placed, hand-animated, tightly controlled worlds. Aggressive LOD streaming, bespoke animations, carefully managed NPC density. Beautiful, cinematic, emotionally resonant. The trade-off is shallow interactivity: you can't pick up a bottle off a table, NPCs don't have real schedules or homes, and the world state barely changes.
Bethesda's Approach
Deeply systemic worlds where every object is tracked, NPCs have routines, and the world state is mutable.
Click to see the trade-offThe Trade-Off
Heavily systemic. Every object is physics-enabled and tracked. NPCs have daily routines, ownership flags, faction relationships. The world state is mutable and persistent. The trade-off is that this systemic overhead is enormous, resulting in games that look a generation behind and cities that feel sparse.
Players have been asking for the convergence of these two approaches for over a decade.
What people want and have never gotten: A world they can live in, that feels both real and nuanced up close, and vast and persistent far away. With real causality and effect cascades. Impact of decisions immediately and over distance and time.
Section 02Why Nobody Has Built This
The incumbents are structurally unable or unwilling to converge on the middle, and the middle is exactly where the biggest unmet demand lives.
The Incumbents Are Structurally Unable
Rockstar has no reason to rebuild RAGE around systemic persistence. Their games print money without it, and adding it would compromise the authorial control that defines their brand. Bethesda has no reason to invest in Rockstar-level visual density and handcrafted feel, because their audience has historically forgiven them for it (until Starfield proved there's a limit). No new studio has the resources to attempt it from scratch using traditional methods.
The Technical Bottlenecks Are Real but Solvable
Projected Save State Complexity
| Stage | Est. Save Size | Est. Load Time |
|---|---|---|
| Early game (10 hrs) | ~60 MB | 15 - 30 seconds |
| Mid game (60 hrs) | ~150 MB | 45 - 90 seconds |
| Late game (150+ hrs) | ~300 MB | 2 - 4 minutes |
| Extreme play (300+ hrs) | ~400+ MB | 5+ minutes |
The Solution Isn't More Power, It's Different Architecture
Modern consoles have the raw horsepower to run such a game. The problem is that nobody has built the engine to make one. It requires:
Hierarchical simulation — full detail near the player, abstracted simulation at distance
Delta-based state persistence — tracking changes from default rather than absolute state for everything
Procedural animation and interaction systems — generating contextual animations on the fly
Aggressive SSD use as virtual RAM for state data, not just asset streaming
A "world janitor" system — NPCs cleaning up, weather degrading objects, plausible narrative tidying
Section 03Feasibility with AI-Assisted Development
A purpose-built engine for one specific vision is a fraction of the work of a general-purpose engine.
Building a competitive AAA engine from scratch traditionally requires 50 to 200 engineers, 4 to 7 years, and $200M to $500M+. That's roughly 2,400 to 16,800 person-months of work.
The AI-Assisted Equation
A purpose-built engine for one specific vision is a fraction of the work of a general-purpose engine.
Scoping the Actual Novel Work
A general-purpose engine that does everything is the 200-engineer project. A purpose-built engine that solves two specific hard problems and does everything else adequately is dramatically smaller.
The Four Novel Engineering Challenges
- Streaming/LOD system tuned to the specific world structure
- Persistent object state architecture — ECS tracking object state as diffs from base world, tiered simulation fidelity by distance
- NPC simulation hierarchy — full AI near player, abstracted schedules at distance
- Interaction/animation system that makes touching objects feel authored rather than janky
Each is 2 to 6 months of concentrated effort with current AI assistance, trending shorter with model improvements.
Estimated Timeline
12 to 24 months of focused work with three coordinated Claude instances to produce a functional engine demonstrating both qualities at a credible level. Not a shipped AAA game, but a vertical slice proving the architecture works: a world section with Rockstar-level visual density where you can pick up every object, NPCs run schedules, and the state persists across sessions.
The vertical slice is everything. Once it works, scaling to a full game is a content and iteration problem rather than an architecture problem. Content production parallelizes well across AI instances.
Section 04The Game
The game trusts the player to be curious, and rewards them for it in proportion to their effort, without ever telling them what to be curious about.
Design Philosophy
The game trusts the player to be curious, and rewards them for it in proportion to their effort, without ever telling them what to be curious about.
The game is simultaneously deeply simulated and deeply authored, where the authored content is hidden inside the simulation rather than sitting on top of it.
Tonal Design: Life, Not Consistency
The game as a whole is intentionally tonally varied. Tonal consistency is an artifact of studios targeting specific vibes for specific crowds. This game rejects that by allowing the player's choices to dictate tone.
Story Architecture: Emergent from Behavior
Stories aren't quests you pick up. They emerge from behavioral patterns intersecting with authored potential.
Every bar, inn, outdoor market contains a handful of potential storylines as deep as the Skooma Stories concept. These storylines don't reveal themselves until the player's actions naturally lead to them. Spending every night in the same bar brings you close to certain characters and groups based on who you talk to and what you say. You can't be friends with everyone, because there's always someone who doesn't like everyone. Most players will never experience most storylines.
The Geology System
The geology of the game world is load-bearing for culture, economy, magic systems, and player progression. It is not a minigame bolted on.
Mining and Synthesis
The Fossil and Paleontology System
Fossils include scans of real fossils, imagined ancestors of mythical species present in the game, and extinct species from the game's own evolutionary history, not from our world or any existing lore.
Identification requires actual scientific method: study the skeletal structure of modern species in-game, compare with partial fossil specimens, work with incomplete data and genuine ambiguity. No quest markers, no field guides, no instructions.
Literature as Unreliable Evidence
Advanced Possibilities
The Skooma Stories Precedent
The game never locks the door. Relapse is always one choice away.
This level of depth, applied to the storylines accessible at every social hub, is the standard. Not every story is about addiction, but every story has this degree of mechanical and emotional integration.
Read the full Skooma Stories document →
Broader Systemic Features
Astronomy
Real sky systems, potentially informed by actual astrophotography
Ecology
Interconnected ecosystems that respond to player actions over time
Economics
Regional economies shaped by geology, trade routes, and player disruption
Language & Scholarship
Dead languages, ancient texts, archaeological discovery
Aging & Mortality
Characters grow old, lose physical abilities, gain others; eventual death
Legacy Systems
Properties, documents, research, relationships carry forward across characters
Social Mechanics
Quantified relationships, reputation systems, NPC memory of player behavior patterns
Section 05The Production Model
Not 100 senior developers. Potentially 1,000 to 5,000 non-developer regular people who care deeply about something specific.
The game is not built by a traditional studio. It's built by one developer with a vision, an AI coordination system, and a distributed network of passionate contributors who are mostly not developers. The bottleneck in engine development isn't raw intelligence. It's coordination overhead and institutional memory. An AI system with persistent memory sidesteps all of that.
The AI Infrastructure
The Contributor Model
Not 100 senior developers. Potentially 1,000 to 5,000 non-developer regular people who care deeply about something specific.
Open-source contribution models work for code because code is modular with clear interfaces. Creative work usually isn't and doesn't. The AI system acts as the translator between "passionate amateur who knows everything about trilobite morphology" and "game engine that needs specific assets, data structures, and behavioral rules." The astronomy person doesn't need to learn UV mapping. They talk about astronomy, provide reference material, and give feedback. The AI handles implementation translation.
Compensation Structure
| Tier | Duration | Benefit |
|---|---|---|
| 0 | < 3 months | Base percent |
| 1 | 3 months | Increased share |
| 2 | 6 months | Further increase |
| 3 | 12 months | Further increase |
| 4 | 24 months | Further increase |
| 5 | 36 months | Maximum share |
Quality Assurance
Design consistency in mechanics and visual features is solvable through iteration, AI coherence, and editorial oversight. May be fragmented early but smoothable over time.
Tonal consistency is not a problem. Varied contributor passions don't create inconsistency; they create life. What might look inconsistent from the outside reads as meaningful depth from the inside, because each contributor's passion comes through in their domain.
Experiential quality testing by domain experts and random playtesters, not just the founder. Set someone up on a specific scenario, have them play through, gather feedback from geologists, aesthetics people, general playtesters.
Call of Duty Zombies: one developer's side passion project became more beloved than the main game for two full titles. Portal: one person's passion for a puzzle mechanic grew into Portal 2. Stardew Valley: pure solo passion project, zero certainty, became the 17th best-selling game of all time.
Section 06Why This Specific Founder
The throughline across all work is intrinsic motivation as an organizing principle. For learning, for play, for commerce, for collaboration.
The game requires someone who thinks about learning and discovery in a specific way, combined with the technical ability and design sense to execute it, combined with AI infrastructure capable of bridging vision and implementation at scale. This particular combination has not existed before.
Educational Philosophy Applied to Game Design
Four years of classroom teaching experience (K-12, mostly 4th through 8th). Neuroscience and neuropsychology background. Founded Roots of Reason: tutoring service built on Self-Determination Theory and intrinsic motivation. Demonstrated ability to identify what drives individual curiosity and create conditions for discovery rather than deliver instructions. The game's design philosophy is Roots of Reason applied to play.
Domain Expertise in Key Content Areas
Founded Lithorium: crystal and mineral business combining geological education with original mythology (Iridescia). Deep knowledge of mineralogy, geology, specimen identification. Understanding of fossil identification and paleontological methods. Natural intersection between educational content design and game world-building.
Technical Infrastructure
Custom AI multi-agent system built from neuroscience fundamentals. 100+ specialized roles, persistent memory systems, communication analysis. Demonstrated ability to produce $8,000 to $15,000 equivalent web development work at a fraction of the cost and time. System improves with every project completed, integrating knowledge across domains.
Human Infrastructure
Proven ability to absorb others' passions and serve their visions. Natural mediator and communicator across varied styles. Instinct for identifying passionate people and creating conditions for their best work. Teaching background means experience translating between experts and non-experts.
The throughline across all work is intrinsic motivation as an organizing principle. For learning, for play, for commerce, for collaboration. It's all one idea: making things that mean something, to someone, and sharing passion in a way that's contagious.
Section 07Valuation Scenarios
Games that deliver on "I can live in this world" don't stop selling. Skyrim is 14 years old. Stardew is nearly 10.
Comparable Data Points
Stardew Valley (solo passion project, farming sim): 50 million copies sold, ~$580M gross revenue, ~$170M estimated net to the developer. No microtransactions, no paid DLC, no corporate backing. Still selling 65,000 units per week nearly a decade after release.
Baldur's Gate 3 (mid-size studio, closest gameplay comparison): 20+ million copies sold by end of 2025, estimated $1.5B+ gross revenue on Steam alone. Larian's 2023 pre-tax profit was €249 million, a twentyfold increase from the prior year. Development cost ~$100M.
Skyrim (the game this most directly competes with conceptually): 60+ million copies sold over 14 years across multiple re-releases. Still selling. The game that defined the "I can just live here" open-world genre.
Scenario Modeling
Comparable to a well-received indie RPG. Even at this floor, the engine IP and AI orchestration system have independent value as licensable technology.
The BG3 trajectory. Delivers on the hybrid promise well enough that word-of-mouth drives sustained sales. Engine licensing becomes viable. Community platform generates recurring revenue.
The Skyrim/RDR2 outcome. Becomes the reference point for a new genre. Sells for decades. Engine becomes the standard for persistent open-world games. The long tail matters enormously here.
The long tail matters enormously here. Games that deliver on "I can live in this world" don't stop selling. Skyrim is 14 years old. Stardew is nearly 10. The genre-defining versions of this concept would sell for decades.
The Engine as Independent Asset
Combined Scenario Summary
| Asset | Conservative | Moderate | Breakout |
|---|---|---|---|
| Game revenue (lifetime) | $15 - 40M | $80 - 250M | $500M - 1.2B |
| Engine licensing (5 yr) | $2 - 10M | $10 - 50M | $50 - 200M+ |
| Orchestration system | $1 - 5M | $10 - 50M | $50 - 200M+ |
| Community platform | $0 - 1M | $5 - 20M | $20 - 100M+ |
| Total potential value | $18 - 56M | $105 - 370M | $620M - 1.7B |
Appendix AQuestions and Answers
Will passion contributors produce quality content?
Yes. Enderal (a total conversion mod for Skyrim built by passionate volunteers) won awards and is widely considered better than most commercial RPGs. The Forgotten City started as a Skyrim mod by one passionate creator and became a standalone commercial hit. Passion contributors with domain expertise, given proper coordination and quality filtering, consistently produce work that rivals or exceeds professional studio output in their specific areas.
Will players respond to this kind of game?
Baldur's Gate 3 sold 20M+ copies by delivering deep systemic interactivity with high production value, and it's not even open-world. RDR2 sold 60M+ copies on authored density alone. Skyrim is still selling 14 years later on exploration and freedom alone. Each of these games delivers part of what this game promises. Nothing delivers all of it.
Can a small team compete with AAA?
Stardew Valley: one person, 4.5 years, 50 million copies, tied with Call of Duty: Modern Warfare in lifetime sales. Among Us: three-person team, became one of the most-played games in history. Dwarf Fortress: two brothers, decades of development, commercial release became an instant hit. The pattern is clear: a focused vision with genuine depth, executed with passion, routinely outperforms hundred-million-dollar productions that lack those qualities.
Won't many contributors produce design inconsistency?
Design inconsistency (mechanical, visual) is real but solvable through iteration, AI coherence checking, and editorial oversight. It's a smoothing problem, not a structural one. Tonal inconsistency, however, is not a risk but a feature. The game's structure (finite lifespan, closing doors, natural behavioral depth) forces tonal consistency at the player level without requiring it at the world level.
Can coordination scale to hundreds or thousands of contributors?
The AI system is already designed for this. Each contributor gets their own instance with its own memory; all data aggregates into a master database. Communication styles are analyzed and adapted to. The system was built from the ground up for multi-agent coordination, not retrofitted onto it.
Appendix CWhat Doesn't Exist Yet
No existing game combines open-world scale with both visual density and systemic persistence, authored emotional depth hidden inside emergent systems, real scientific methods applied to fictional data, finite character lifespan creating natural tonal consistency, generational play with persistent world state, storylines that emerge from behavioral patterns rather than quest pickups, and a world where contributor passion is the content pipeline.
Dwarf Fortress
Deepest simulation
The gapZero authored feel
Disco Elysium
Narrative depth, substance mechanics
The gapNot open-world, not persistent
Kenshi
Player agency, indifferent world
The gapNo authored narrative
Baldur's Gate 3
Consequence, character depth
The gapNot open-world, not persistent
No Man's Sky
Vastness, procedural generation
The gapShallow, no authored stories
Skyrim
Open world, exploration, quantity
The gapVisual density, NPC depth
RDR2
Visual density, authored feel
The gapShallow interactivity, no persistence
Stardew Valley
Passion-driven, deep systems
The gapNot 3D open-world scale
Appendix DThe Underlying Philosophy
"Passion is nothing alone, it's something to be shared."
The game, the engine, the production model, and the founder's existing businesses are all expressions of a single principle: intrinsic motivation as an organizing force.
- Roots of Reason: Kids learn when curiosity leads them somewhere real.
- The game: Players experience meaning when curiosity leads them somewhere real.
- The production model: People do their best work when their passion is genuinely valued.
- Lithorium: A piece of basalt, seen through the eyes of 40 curious fifth graders, became a museum, a mythology, and a business.
The game is the medium large enough to hold all of it at once. Its purpose is not to be a game. Its purpose is to fund and prove the philosophy that drives everything else. A world where passionate work, followed through, persists.