'We Do Not Let You Into Every Building for Your Own Good!' — No Law's Dense Cybergrunge City Might Just Be the Most Alive Open World Ever Created

Introduction: When Every Door Is a Risk Worth Taking

Open worlds are sales pitches. From the neon-soaked favelas of Cyberpunk 2077 to the sprawling wastes of Grand Theft Auto VI, these cities beg you to explore-but they also lie. They promise infinite possibility, then lock 99% of their interiors behind loading screens or inaccessible facades. No Law's Cybergrunge City doesn't just open the doors; it dares you to enter, knowing that every threshold might be your last. In an exclusive interview with Neon Giant, the Swedish studio that brought us The Ascent, we learned why their upcoming title might redefine what it means for a virtual world to feel truly alive.

When lead designer Johannes Lindqvist told me, "We don't let you into every building for your own good," it wasn't a threat-it was a design manifesto. In No Law, every accessible interior serves a purpose: a story - a trap, a reward. Or a lesson. But more importantly, the city's density isn't cosmetic. It's engineered using techniques that borrow as much from AI state machines and procedural generation as from traditional level design.

The Philosophy of 'No Law': Why Every Building Matters

The game's title isn't merely a branding choice. In development circles, "no law" refers to a design constraint system where rules are emergent, not enforced. Lindqvist explains, "We started with the question: What if an open world didn't have invisible walls, but instead used consequence as its barrier? " The result is a city where entering a building marked "Private" might trigger a lethal defense drone. Or it might open a hidden bazaar. The uncertainty creates a tension that traditional fast-travel economies can't replicate.

This approach forced Neon Giant to rethink their tech stack. Instead of pre-baking every interior as a separate level, they developed a modular system that combines handcrafted rooms with procedurally assembled layouts. "We use a Poisson disc sampling algorithm to distribute points of interest," Lindqvist says, "then a custom behavior tree dictates what each interior 'wants' to be-a cache, an ambush, a side quest hub. " The result is a city where no two playthroughs explore the same sequence of spaces.

Procedural Density Meets Handcrafted Chaos

Creating a believable urban environment at this scale requires a marriage of two adversarial techniques: procedural generation for volume. And hand-authoring for character. Neon Giant's city is built on a layered system that assigns each block a "biome" not of foliage. But of functional density. For example, the "Cybergrunge" district uses a combination of Voronoi diagrams to define gang territories Unity's GameObject instantiation pipeline to scatter thousands of interactive props.

In production, we found that the hardest part wasn't generating the geometry-it was ensuring that every generated space could be navigated intelligently by the player. "Our NavMesh generation hit a wall at around 50,000 interior cells," recalls technical director Elin Söderström. "We had to implement a hierarchical pathfinding system that caches connectivity graphs for each district, then stitches them together at runtime. " This mirrors techniques used in large-scale simulations like Microsoft Flight Simulator's bing map integration. Where data must be chunked and prioritized without destroying immersion.

AI That Remembers: The Living City's NPC Ecosystem

The most impressive aspect of Cybergrunge City's aliveness is its non-player character (NPC) ecosystem. Every citizen has a daily schedule, but unlike The Elder Scrolls games, schedules break when the player intervenes. "We store state for each NPC across a 7-day cycle using a time-series database compressed into a 64-bit integer," Söderström explains. "When you rob a store, the clerk remembers you for three in-game days and will call security on sight. " This isn't a simple flag-it's a reputation system that propagates through social networks using a graph database under the hood.

During our interview, the team demonstrated a scenario where a player repeatedly broke into apartments in a specific district. After the fifth break-in, the entire block's NPCs began locking their doors-not as a scripted event. But because the game's AI used a reinforcement learning variant (Q-learning with state hashing) to adjust behavior patterns. "We're not simulating a thousand individual brains," Lindqvist says. "We're simulating a thousand tiny decision trees that share a common reward signal: survival. "

Performance at Scale: Rendering a Million Interiors

How do you render a city where every building can be entered without loading screens? The answer is a hybrid of streaming tech and Level-of-Detail (LOD) hierarchies that would make the engineers behind Vulkan compute shaders proudNeon Giant implemented a custom frustum culling system that doesn't just cull geometry-it culls entire interior volumes based on their distance from the player and the probability of intrusion.

"We use a spatial hash grid where each cell stores a bitmask of which buildings are visible from the player's current position," says Söderström. "When you're on the street, only the first floor's interior geometry is loaded. When you approach a door, the system predicts which rooms you're likely to enter and loads those at full detail. " This predictive loading leverages a Markov chain model trained on telemetry from beta testers. In our testing on a mid-range RTX 3060, frame times stayed below 16ms even when walking through the densest market district-a feat that required batching draw calls using GPU-driven culling.

The underlying engine is an in-house modification of Unity's DOTS (Data-Oriented Tech Stack), extended to handle what the team calls "portal chains. " Each interior is a portal to potentially hundreds more, forming a graph that the renderer must traverse in real time. This is conceptually similar to the RSM (Reflective Shadow Map) techniques used in CryEngine. But adapted for interior-exterior transitions.

The Designer's Dilemma: Balancing Exploration and Safety

If every door is a risk, how do you prevent player frustration? This is where Neon Giant's years of experience with The Ascent paid off. "In The Ascent, players could accidentally trigger a massive firefight by walking into the wrong room," Lindqvist recalls. "We learned to telegraph danger through environmental storytelling-flickering lights, bloody footprints, or the sound of a security drone powering up. " In No Law, these cues are procedurally generated based on the room's tag (e g., "gang hideout" rooms always emit low-frequency hums and a slight red tint to the lighting).

Another crucial lesson was to avoid punishing players for curiosity. Every dangerous interior also contains a reward that mathematically justifies the risk. The team uses a Monte Carlo simulation to compute expected value per interior type, then adjusts loot tables dynamically. "If 90% of players die in a certain type of building, we increase the loot quality by 1. 5x and add a visible health pack just inside the door," says Lindqvist. This data-driven balancing is run offline between patches. But the final values are baked into asset bundles to avoid runtime overhead.

From Neon Giant's Past: Lessons from The Ascent

Neon Giant's previous title, The Ascent, was a top-down shooter set in a cyberpunk world. It earned praise for its dense, lived-in environments. But also criticism for repetitive interiors and lack of verticality. No Law is their answer to those criticisms. "We took the procedural tile-based system from The Ascent and rebuilt it for first-person," Söderström explains. "Instead of 2D grid rooms, we generate 3D volumes with multiple floors, connected by stairs and elevators that are themselves procedurally placed. "

The team also applied lessons from the LOD (Level of Detail) research that made The Ascent run on last-gen consoles. For No Law, they extended the concept to AI updates: NPCs at long distances use a simplified behavior tree with only 5 states (idle, patrol, flee, alert, fight) instead of the full 20-state array used for close interactions. This reduces CPU load by 70% in crowded scenes.

The Future of Open Worlds: What Others Can Learn from No Law

No Law's innovation isn't just technical-it's philosophical. By refusing to let players into every building without consequence, Neon Giant has created an open world that respects player agency more deeply than those that offer infinite but shallow access. The "dense cybergrunge" aesthetic isn't a visual gimmick; it's a direct result of the underlying simulation. Every neon sign, every flickering hologram. And every locked door is a node in a vast graph of cause and effect.

For other developers, the lesson is clear: treat every door as a design decision, not a technical limitation. "We started by asking what our players would miss if they never saw an interior," says Lindqvist. "The answer was nothing-because the city's life is in the cracks, not the facades. " This mindset, combined with the engineering rigor of hierarchical pathfinding and reinforcement learning NPCs, could inspire a new generation of world-building tools. Imagine a city generator that outputs not just buildings. But also the rules for how those buildings interact with the player's history, and that is the promise of No Law

Frequently Asked Questions

Conclusion: A Living City Worth the Risk

No Law isn't just another cyberpunk game. It's a case study in how to make an open world that respects player time, curiosity. And intelligence. By grounding its design in real-time systems engineering-branch prediction for AI, hash-based spatial grids. And reinforcement learning behavior-Neon Giant has delivered a world that feels as unpredictable and responsive as a real metropolis.

If you're a developer, take note: the next generation of open worlds will be judged by what they allow you to touch, not what they show you. If you're a player, prepare to enter buildings you were warned about. And if you're anyone else, read our full interview with Neon Giant for deeper dives into the tech. No Law releases on PC and consoles later this year. The doors are open-just don't expect to survive every threshold,

What do you think

Should open-world games prioritize interior density over map size,? Or is the trade-off in rendering complexity not worth the immersion gain?

Do you agree that NPCs with long-term memory make exploration more engaging,? Or do they risk creating overly punitive gameplay loops?

How should developers balance procedural generation with handcrafted content when creating spaces that feel both unique and narratively meaningful?