Will Wright Unveils Spore to Counter Soaring Next-Gen Content Budgets

Will Wright GDC 2005 Spore (The Future of Content) Remastered — Zeusmonomon

Zeusmonomon

Procedural generation and automated player authorship solve the game industry's next-generation budget crisis by turning lightweight player instructions into functional worlds without ballooning studio headcount.

AI-assisted summary of Will Wright GDC 2005 Spore (The Future of Content) Remastered — original source. Check the original for full context.

Source published:

Key ideas

  • Will Wright unveiled the prototype of Spore at GDC 2005, directly challenging the industry consensus that next-generation HD gaming would trigger an unsustainable asset production crisis for developers. (0:00)
  • Spore circumvents massive storage requirements through procedural generation, storing complex user creations as ~1 kilobyte parameter sets that expand at runtime into ~5 megabyte assets—achieving a 5,000-to-1 compression ratio. (26:35)
  • The game operates on an asynchronous multiverse model where player-authored organisms, buildings, and vehicles are dynamically downloaded from a central server to populate other players' ecosystems based on parametric queries. (14:35)
  • Spore combines the exponential macroscopic progression of Charles and Ray Eames' Powers of Ten with SETI's Drake equation, transitioning players from cellular life up to galactic civilization. (58:27)

Detailed notes

## Background Published on YouTube by channel Zeusmonomon on August 27, 2020, this video is a remastered recording of Will Wright's landmark presentation at the 2005 Game Developers Conference (GDC). Titled 'The Future of Content', the talk served as the surprise industry reveal and live technical demonstration of his ambitious evolutionary project, Spore.

## TL;DR Addressing industry anxieties over ballooning next-generation asset production costs, Will Wright argues that procedural generation and player agency offer a scalable path forward by turning consumers into world authors. Demonstrating an early alpha prototype of Spore, he showcases seamless progression across astronomical scales—from a 2D cellular tide pool to terrestrial creature evolution, municipal architecture, and full interstellar exploration. Wright details how procedural inverse kinematics, behavioral synthesis, and automated 'artist in a box' pipelines achieve a 5,000-to-1 compression ratio, using lightweight mathematical instruction sets (~1 KB) to asynchronously populate a shared galactic ecosystem.

## Atmosphere & style The presentation carries the electrifying atmosphere of an unexpected paradigm shift, blending Wright's deadpan, self-effacing wit with ambitious technological evangelism. Watching the talk feels like being in a hushed keynote hall witnessing groundbreaking computational magic held together by 'bailing wire and duct tape,' where the audience oscillates between stunned laughter and awe-inspired applause. Wright's delivery is casual yet deeply erudite, casually bridging references from the demoscene and Care Bears lore to evolutionary biology and the Drake equation with infectious creative wonder.

## Bulletin - [00:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=0s) Will Wright unveiled the prototype of Spore at GDC 2005, directly challenging the industry consensus that next-generation HD gaming would trigger an unsustainable asset production crisis for developers. - [26:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1595s) Spore circumvents massive storage requirements through procedural generation, storing complex user creations as ~1 kilobyte parameter sets that expand at runtime into ~5 megabyte assets—achieving a 5,000-to-1 compression ratio. - [14:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=875s) The game operates on an asynchronous multiverse model where player-authored organisms, buildings, and vehicles are dynamically downloaded from a central server to populate other players' ecosystems based on parametric queries. - [58:27](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3507s) Spore combines the exponential macroscopic progression of Charles and Ray Eames' Powers of Ten with SETI's Drake equation, transitioning players from cellular life up to galactic civilization.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=0s) Next-generation HD game development risks squeezing out smaller studios due to escalating asset production costs and diminishing returns on visual content creation. - [03:44](https://www.youtube.com/watch?v=ofA6YWVTURU&t=224s) Adding player agency and ownership creates a third dimension of perceived value (the Z-axis), multiplying game depth without linearly increasing art budgets. - [05:24](https://www.youtube.com/watch?v=ofA6YWVTURU&t=324s) Emergent player-driven narratives, such as idiosyncratic roleplaying in Grand Theft Auto, resonate far more deeply with players than heavily scripted, authored cutscenes. - [06:41](https://www.youtube.com/watch?v=ofA6YWVTURU&t=401s) The advent of the 650 MB CD-ROM shifted the industry ratio from code-and-algorithm reliance on early microcomputers toward massive pre-baked data storage. - [08:30](https://www.youtube.com/watch?v=ofA6YWVTURU&t=510s) Techniques adapted from the demoscene demonstrated that procedural algorithms can compress vast visual and auditory experiences into miniscule binaries. - [15:52](https://www.youtube.com/watch?v=ofA6YWVTURU&t=952s) All creature locomotion and physical interactions in Spore are generated on the fly via procedural inverse kinematics rather than hard-coded animation clips. - [21:36](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1296s) Unconventional creature anatomies grant procedural animation creative leeway, as players lack rigid real-world expectations for how alien forms should move. - [26:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1595s) Lead artist Ocean Quigley's 'artist in a box' concept automates complex meshing, UV unwrapping, texturing, and skeletal rigging behind an intuitive player interface. - [28:39](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1719s) Spore's offline distribution media included tens of thousands of procedural assets pre-generated using the identical algorithmic sliders given to players. - [39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) The game intentionally defies standard industry rules against genre mixing, integrating distilled mechanics from cellular action, creature survival, city building, and 4X space strategy. - [39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) Creative level editors redefine player identity away from being a protagonist like Luke Skywalker or Frodo Baggins to becoming an author like George Lucas or J. R. R. Tolkien. - [47:43](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2863s) Terraforming in the space stage takes inspiration from Kid Pix, prioritizing tactile, chaotic, and unpredictably playful tools over rigid scientific modeling. - [55:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3321s) Spore employs a T-shaped design architecture where linear, evolutionary stages serve as tutorials before unlocking an open vertical and horizontal galactic sandbox. - [59:47](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3587s) Over 200 experimental prototypes were developed by Maxis to test and validate Spore's untried procedural systems prior to full production. - [01:00:54](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3654s) Wright concluded that the primary bottleneck in groundbreaking game design is often internal self-doubt rather than external publisher or technical resistance.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=0s) Introduction and the "Next-Gen Content Crisis"

After a brief introduction and warm applause from the Game Developers Conference audience, Will Wright takes the stage, joking about his booming microphone sounding like “the voice of god.” Before diving in, he handles the standard conference housekeeping—instructing attendees to switch off cell phones, refrain from taking photographs, and remember their evaluation forms, dryly adding, “and they made me say all that stuff.” He introduces Chris, a co-designer who has worked alongside him for roughly five years and will be assisting him with the presentation. Wright opens with a quick apology to the simultaneous translators in the room, warning them that he talks notoriously fast when trying to cram an enormous amount of material into a single hour: “If you can't keep up, feel free to just make up stuff, and it might be better than what I'm saying.”

Turning to the session’s official topic, Wright admits that its listing, “The Future of Content,” is hopelessly vague. Typically, GDC organizers call him months in advance to ask for a presentation topic; he usually spends a week pondering general concepts before handing over a generic working title. This year, however, was different: “When they asked me what I was going to talk about, I just flat out lied.” While he will touch on content, his true intent is to challenge the prevailing anxiety rippling through the game industry regarding the looming transition to high-definition, next-generation hardware. The conventional wisdom across the conference halls suggests that next-generation games will demand unprecedented, unsustainable labor costs that will choke out innovation.

Will Wright outlines this looming crisis and his proposed alternative: “It seems like the theme that I'm coming away from this conference with so far is that the next generation titles are going to be incredibly expensive, we're going to have to build mountains of content, and it's going to drive most of the smaller players out of the market.” Rather than accepting this brute-force arms race of asset production as an inevitability, Wright argues that the industry needs to rethink its approach to world-building entirely.

[02:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=124s) Diminishing Returns and the Power of Player-Generated Content

Proposing that an alternative title for the session could be “What I Learned About Content from *The Sims*,” Wright examines the explosive, non-linear growth of development overhead. Displaying the content inventory for *The Sims 2*, he highlights the sheer mountain of labor required compared to the original game. As development teams expand across console generations, the proportion of personnel dedicated exclusively to content creation grows disproportionately fast compared to the rest of the team. This creates an economic trap governed by basic utility theory: building twice as much content takes twice as much human effort, yet it fails to deliver double the perceived value to the user.

Wright illustrates this principle through a desert island metaphor: if an individual is starving on an island, a single apple holds life-saving utility, but “once you have 100 apples, you know, the 101st apple is not worth that much.” Developers are pouring exponential resources into crafting additional assets that yield diminishing emotional and financial returns. In sharp contrast, players demonstrate an insatiable enthusiasm for creating their own content whenever they are given the appropriate toolsets. Even accounting for varying levels of amateur craftsmanship, the sheer volume of high-quality community output vastly outstrips what an internal development studio could ever hope to produce manually.

- *The Sims 2* official exchange had already accumulated approximately 17,000 player-uploaded characters and 10,000 houses shortly after launch. - Even if only 5% to 10% of user submissions meet professional standards, the rapidly improving quality curve yields a massive, free reservoir of engaging content. - A prominent player fan site recently celebrated hosting its 100,000th custom-uploaded object.

[03:44](https://www.youtube.com/watch?v=ofA6YWVTURU&t=224s) The Content Value Equation: Adding the Ownership Dimension

Wright notes that monitoring player telemetry and behavioral metrics in *The Sims* provided critical design insights: tracking which verbs players select, what items they purchase, and how they navigate systems reveals how people actually play. A primary takeaway is that players fundamentally despise arbitrary boundaries, especially unnatural invisible walls that block their curiosity. While the industry is shifting toward more expansive, open-ended environments to address this, widening the possibility space typically demands steep, non-linear iteration cycles to balance emergent interactions.

```mermaid flowchart LR accTitle: Player Value Dimensions accDescr: Diagram illustrating how adding player ownership expands value beyond the traditional 2D model of content quantity versus quality.

subgraph Traditional["Traditional Content Model"] direction TB QNT["Content Quantity<br/>(e.g., Katamari Damacy)"] QLT["Content Quality<br/>(e.g., Pikmin)"] VAL2D["2D Value Area<br/>(Quantity × Quality)"] QNT --- VAL2D QLT --- VAL2D end

subgraph Expanded["Wright's Proposed Model"] direction TB OWN["Player Ownership<br/>& Agency"] VAL3D["3D Value Volume<br/>(Massive Exponential Area)"] VAL2D -->|"Extrude along Z-axis"| OWN OWN --> VAL3D end ```

To evaluate content efficiency, Wright critiques the implicit industry assumption that player value is strictly a two-dimensional calculation multiplying asset quantity by visual fidelity. Games like *Pikmin* feature relatively modest quantities of content polished to high visual fidelity, whereas games like *Katamari Damacy* offer sprawling arrays of low-fidelity objects, yet both land in a comparable sweet spot of overall player value. However, Wright argues this 2D gradient completely ignores a critical third dimension: player agency and ownership. By allowing players to author, customize, and invest themselves into the assets in their world, developers can unlock an enormous volume of value without incurring exorbitant production budgets:

> “I think, you know, the ownership that we can give players over content feels like a very meaningful new dimension that, you know, by pulling that out just a little bit, we add tremendous amount of area to that player value.”

[05:24](https://www.youtube.com/watch?v=ofA6YWVTURU&t=324s) "Moe" from Grand Theft Auto and the Primacy of Player Stories

To illustrate how agency transforms engagement, Wright shares a personal anecdote from playing *Grand Theft Auto: San Andreas*. Instead of pursuing the main narrative missions, he created a custom persona named “Moe,” a total slacker who spent his days wandering the map, learning minor bicycle tricks, and ignoring criminal syndicates. Wright customized Moe to wear nothing but underwear and a Barbie wristwatch. The absurd juxtaposition of serious mob bosses treating this half-naked bicycle slacker with deadly gravity during dramatic story cutscenes became Wright's favorite part of the game.

When he finished his play sessions, Wright found himself eager to share stories with his wife and daughter about Moe’s mundane, off-beat misadventures rather than the scripted plot points authored by the studio. For Wright, this dynamic proves that players form deep emotional attachments to events they initiate themselves, whereas pre-rendered cinematic moments rarely leave lasting, shareable memories. Emphasizing why game design should pivot toward generative autonomy, Will Wright insists: “And so I contend that player stories will always be more powerful than scripted stories that we try to tell the players.”

[06:41](https://www.youtube.com/watch?v=ofA6YWVTURU&t=401s) From Apple II Code to the CD-ROM "Meteor"

Synthesizing these insights—the shift toward non-linear agency, spiraling art costs, the power of player-authored content, and behavioral analytics—Wright arrives at the true thesis of his presentation: “What I learned about content from *The Sims* and why it's driven me to procedural methods.” To explain how game development arrived at its current predicament, he reflects on his early career developing for the Apple II, a machine an original advertisement jokingly touted as being six times faster than a teletype.

In the early personal computer era, developers worked down to the bare metal in assembly and machine code, keeping explicit physical memory maps to track every byte. Because 140-kilobyte floppy disks and system RAM were severely restricted, a typical program's memory footprint was split evenly: roughly 50% code and 50% data. In modern games, that ratio has skewed drastically to 1 to 100 in favor of pre-baked assets.

Wright identifies the pivotal historical turning point as the arrival of the 650-megabyte CD-ROM, an optical format that looked like an unfathomable supertanker compared to 5.25-inch floppies: “Well, actually, before I talk about that, I mean, this basically was kind of like the meteor that came in and sounded the death knell for the algorithm.”

Data immediately ascended to supremacy. Wright points to Cyan’s *Myst*, created by Robyn and Rand Miller in HyperCard, as the definitive triumph of the data-heavy era. While *Myst* was functionally an elaborate prerendered slideshow, Wright found it astonishingly immersive and brilliantly designed. Even so, playing it left him with an inescapable realization: the designers had experienced all the true fun of building the world, leaving players to passively click through the static aftermath of that creation.

[08:30](https://www.youtube.com/watch?v=ofA6YWVTURU&t=510s) The Demoscene and the Power of the Algorithm

While the mainstream games industry pursued data-driven pipelines—stuffing CD-ROMs with pre-rendered, hand-crafted assets—a different movement kept an older tradition alive. Will Wright describes this subculture as a separate tribe centered in Northern Europe, Scandinavia, and specifically Helsinki: the demoscene. Wright observes: “They basically worshipped the algorithm, kind of kept practicing their black arts, they adopted kind of modern dress and technology, but... we call it the demoscene, and these guys are basically crazy.”

```mermaid flowchart LR accTitle: Demoscene Data Compression Contrast accDescr: Visualizing the massive mathematical scale difference between 64 kilobytes and a full CD-ROM. subgraph Demo["Demoscene"] A["64 KB Executable"] end subgraph CD["Mainstream CD-ROM"] B["(64 KB)^2"] --> C["Squaring Steps..."] --> D["650 MB CD-ROM Capacity"] end A -.->|"Algorithmically produces equivalent visuals"| D ```

These programmers trace their origins back to software cracking groups who attached animated cracktros to showcase their technical prowess. Over time, these grew into elaborate competitive art pieces where developers pack massive procedural animations, visual effects, and music into binaries as compact as 64 kilobytes. To illustrate the magnitude of difference between algorithmic generation and raw data storage, Wright notes that if you take 64 kilobytes and repeatedly square it, you arrive at the capacity of a standard CD-ROM.

- Demoscene origins lie in competitive software cracking groups signing their releases with intro screens. - Programmers rely on procedural generation to fit expansive audio, visual, and animated routines into 64-kilobyte files. - Wright found this algorithmic philosophy a powerful alternative to the industry's ballooning content budgets.

---

[10:18](https://www.youtube.com/watch?v=ofA6YWVTURU&t=618s) The Microscope Tangent and Assembling the Team

Reflecting on his experiences developing *The Sims*, Wright reframes the thesis of his keynote: “What I learned about content from *The Sims*, why it’s driven me to procedural methods, and what I now plan to do with them.” To bring this vision to life, he recruited an elite team skilled in procedural programming and set them up in a covert development studio.

Before revealing the prototype, Wright shares a parenting tangent from when his daughter was five years old. He had set out to buy both a high-end telescope and a quality microscope. While telescopes were widely available at retail science shops like The Nature Company, finding a good stereoscopic microscope proved nearly impossible until online marketplaces like eBay emerged. He advises the audience on the best hobbyist setup: buy a low-power stereo microscope equipped with overhead illumination. Wright explains that while high magnification seems intuitive, low magnification produces a revelatory experience: “You’ll put your hand under this thing and you’ll spend an hour just obsessing over your hand, and it’s like a religious experience.”

- Wright opted to unveil the prototype at the Game Developers Conference rather than a commercial marketing venue like E3 due to his kinship with fellow developers. - The demonstration build was in an early alpha state, described by Wright as being “held together with, you know, bailing wire and duct tape.”

---

[12:41](https://www.youtube.com/watch?v=ofA6YWVTURU&t=761s) The Cellular Phase: Procedural Pac-Man

Moving over to the display screen, Wright presents the prototype, titled *Spore*. The demonstration begins at the micro scale, placing the player in control of a single-celled organism drifting through a fluid-dynamic environment. Wright characterizes the phase as “a very analog, fluid version of Pac-Man.” The player maneuvers the microbe around hazardous elements, hunts floating nutrient pellets, and evades hostile organisms that can kill the player's creature on contact.

```mermaid flowchart TD accTitle: Cellular Lifecycle Loop accDescr: The loop of consuming nutrients, laying an egg, editing creature morphology, and returning to the game. A["Guide Cell & Eat Nutrients"] --> B["Lay Egg"] B --> C["Enter Cell Editor"] C --> D["Attach Parts: Mouths, Locomotion, Weapons"] D --> E["Hatch Modified Organism"] E --> A ```

Consuming enough food allows the organism to reproduce by laying a small yellow egg. Clicking on this egg opens an editor where the player can evolve the microbe using modular biological components. These parts fall into distinct functional categories: mouthpieces (such as filter feeders or piercing proboscises that dictate dietary habits), locomotion appendages (cilia, flagella, or micro-jets), and defensive weapons like spikes or nematocysts. Wright attaches a defensive spike to the front of his cell, returning to the simulation with new offensive capabilities that transform former predators into viable food sources.

- The entire cellular habitat and its fluid dynamics are procedural. - Part attachments alter core game verbs, expanding diet options, locomotion speed, and combat strength. - As the creature grows, the camera gradually pulls back, shifting depth of field to reveal larger layers of the microscopic ecosystem.

---

[14:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=875s) The Tidepool Phase and the Asynchronous Multiverse

Skipping forward several evolutionary cycles, the game transitions from a two-dimensional cellular broth into a three-dimensional aquatic environment beneath the ocean surface. Here, player creation integrates directly with procedural systems: players sculpt the morphologies, while procedural animation, behavior trees, and surface textures are synthesized on the fly.

Crucially, Wright reveals that the creatures inhabiting the local world are not authored by Maxis designers or spawned purely at random; they are pulled asynchronously from other players over the network:

> “When I say coming from other players, what I mean is that they are asynchronously being sent to your computer to fill out your world. We’re actually building a balanced ecosystem depending on what your world needs; we deliver other players’ content to pull out that world around you.”

- The transition shifts the simulation from micro-scale 2D fluid dynamics to a full 3D marine environment. - The procedural ecosystem balances its food webs by dynamically fetching user-created organisms suited to specific ecological niches.

---

[15:52](https://www.youtube.com/watch?v=ofA6YWVTURU&t=952s) The Creature Editor and the Emergence of Land Animals

When the marine organism reproduces, the player enters the creature editor. Wright describes the editing suite as a creative hybrid of “Mr. Potato Head, Erector Set, and clay.” The player directly manipulates an underlying spine and skeletal structure that dictates body topology. Scrolling the mouse wheel inflates or narrows flesh around the vertebrae to sculpt the torso, while modular body components provide functional traits:

| Part Category | Biological Implementations | Primary and Secondary Functions | |:--- |:--- |:--- | | **Graspers** | Claws, Hands, Pincers | Object manipulation, secondary attack utility | | **Locomotion** | Fins, Multi-jointed legs, Cilia | Swimming, terrestrial walking, speed modification | | **Appendages** | Tails, Spines, Feet | Balance, defense, swimming propulsion |

Wright strips the fins off his aquatic creature, drastically diminishing its swimming performance, and replaces them with a three-jointed leg and foot assembly to construct an asymmetrical tripod. Because the creature can no longer navigate the water efficiently, Wright walks the beast up the ocean slope onto dry land.

The software inspects the user-authored bone hierarchy, joint counts, and body weight, and then procedurally calculates locomotion on the fly: “The walking you see here is all procedurally generated based upon an analysis of the skeleton that you created; in fact, every verb in the game is as well.”

- Parts can be scaled, rotated, warped, or removed entirely from the skeletal structure. - Locomotion mechanics and animation cycles are generated purely through dynamic physics and kinematic analysis rather than static, pre-recorded animation clips. - Leaving the water marks the evolutionary leap from the marine phase to terrestrial survival.

[18:22](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1102s) Creature Combat, Procedural Verbs, and Mating

Will Wright guides his newly evolved creature into the wild, explaining that since his creature is a carnivore, survival hinges on hunting down prey. He directs the creature toward an unsuspecting target and attacks. Wright points out that the creature's combat style is directly tied to the anatomy defined in the editor; because he placed the creature's weapon on its tail, it attacks by whipping its rear end around. Positioning matters: if the creature is not close enough, the attack misses entirely. Luckily, Wright's creature has enough speed to chase down the prey, kill it, and begin feeding.

```mermaid flowchart LR accTitle: Procedural Verb Mixing Flowchart accDescr: Diagram illustrating how the Spore engine blends two base actions, biting and walking, into a newly emergent compound action called dragging. subgraph S1["Atomic Player Commands"] A["Command: Take a bite"] B["Command: Walk forward"] end subgraph S2["Procedural Synthesis Engine"] C["Blend Animation & Kinematics"] end subgraph S3["Emergent Game Verb"] D["Result: Drag prey carcass"] end A --> C B --> C C --> D ```

Wright uses this moment to demonstrate the power of procedural verbs, showing how fundamental actions can be cleanly blended by the engine to yield brand-new behaviors. By commanding the creature to take a bite out of the carcass while simultaneously ordering it to walk forward, the system synthesizes an entirely new verb: dragging. He underscores the broader consequence for gameplay:

> “One of the real wins though is that whatever I've created in this game is going to behave differently—I mean, you know, it's just going to be fundamentally different behavior than anything that you might create in the game, which is a huge win.”

Danger suddenly arrives when an aggressive apex predator appears. Wright scrambles to flee, noting that the game automatically populates a balanced ecosystem where the player is not guaranteed to be at the top of the food chain. He outruns the pursuer, hoping the beast will distract itself by devouring a few nearby hopping creatures instead. Once safe, Wright transitions to the life-cycle loop: a creature lives out its lifespan, gathers in-game currency, and performs a mating call to find a partner. Finding a mate and successfully performing a quirky procedural mating dance yields an egg, which grants the player access to the creature editor to evolve the next generation.

[21:36](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1296s) Emergent Morphology: From Buttface to Care Bears

Wright returns to the creature editor to showcase the system's open-ended expressiveness by loading a series of bizarre community and internal builds. He presents a creature nicknamed "Buttface," which features an unconventional, backwards leg arrangement. Despite the bizarre anatomy, the procedural animation system makes it walk smoothly and naturally. Wright points out a psychological trick at play: because players have never seen a creature like Buttface in real life, they have no ingrained expectations of how it ought to walk, giving the animation engine far more leeway than it would have with a humanoid.

He then demonstrates several other creations to show the breadth of the system: - **Tweety Bird**: A creature sporting an enormous head on top of a slender frame, which Wright compares to an unstable SUV like a Ford Explorer, joking that it would probably suffer from very low evolutionary fitness. - **Hang Dog**: A slumped, slouching creature with an exceptionally cool, languid walking cadence. - **Dilemma**: A spindly, elongated creature whose skeletal topology happens to produce an intensely rapid, scuttling gait. - **Care Bear**: A custom-built, cuddly pink bear that Wright insisted the development team include, though he jokes he will turn it into a vicious carnivore.

Wright stresses that every single one of these wildly divergent forms runs on the exact same underlying animation script. The physical performance and movement speed of a creature are not pre-baked stats; they are emergent properties arising from how the procedural inverse kinematics interpret the skeleton and part topology chosen by the player.

[24:01](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1441s) Player Ownership and Pop-Culture Archetypes

Reflecting on past evolution games, Wright argues that most titles made players feel alienated from their creations because players were restricted to choosing from a rigid set of fixed parts or slotting heads, torsos, and tails into predefined sockets. He contrasts this with the creative friction of *The Sims*, where players had to leave the game, download external utility programs, make a custom item, and manually upload it to a fan site. For *Spore*, Wright's core design philosophy is to embed creation tools seamlessly inside the primary gameplay loop and make content sharing entirely transparent: the game should automatically study a player's habits and quietly pull down matching, high-quality creatures made by other players across the world without anyone touching a browser.

Wright draws a parallel between this creative ownership and the way children master pop-culture rule sets:

> “By learning the different types and the names of the dinosaurs, you know, it's kind of cool that I knew what an Ankylosaurus was, my mother didn't, you know, and that gave me a certain ownership of those.”

Whether it was memorizing dinosaur match-ups in his own youth, mastering the elaborate mechanics of *Pokémon*, or collecting *Neopets*, children naturally project deep narratives onto systems with defined traits.

This leads Wright into an extended, comic analysis of the *Care Bears* franchise. Originally conceived by American Greetings as greeting cards with emotional traits and specialized superpowers, they gradually accumulated a complex mythology. Wright shares that an online psychology quiz identified him as Sunshine Bear, then riffs on the lore of the Care Bear Cousins (who are not bears at all) and their cosmology: the Care Bears inhabit Care-a-Lot in the sky, while the Cousins live below them in the Forest of Feelings. When falling from clouds, they pass the forest and descend all the way to Earth for their rescue missions—placing Care-a-Lot roughly in low Earth orbit. Wright quips that while the entire realm is called the Kingdom of Caring, the creators implicitly treat Earth as the "kingdom of who gives a crap," since it is the troubled place in perpetual need of rescue.

[26:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1595s) "Artist in a Box" and Creative Amplification

Returning to technical architecture, Wright introduces a guiding concept coined by the game's lead artist, Ocean Quigley: building an "artist in a box." The development team inventoried every laborious stage of commercial art production—mesh generation, UV unwrapping, texturing, rigging, keyframing, and scripting—and set out to automate them entirely. By placing an automated production pipeline behind a clean set of intuitive controls, the game grants ordinary players immense creative leverage. Wright outlines the game's stylistic spectrum as spanning from the family-friendly whimsy of Pixar (like Dr. Seuss worlds) to the dark biomechanical horror of H.R. Giger.

```mermaid flowchart TD accTitle: The Spore Creative Amplification Pipeline accDescr: Pipeline diagram illustrating how minimal player inputs are expanded into full-fidelity in-game assets via procedural generation. subgraph Input["1. Player Creative Input (~1 KB)"] A["Simple Topology & Part Placement"] end subgraph Generation["2. 'Artist in a Box' Procedural Systems"] B["Polygon Mesh Synthesis"] C["Topology-Aware Procedural Textures"] D["Procedural Rigging & Animation"] E["Autonomous AI & Player Controls"] end subgraph Output["3. Runtime Game Output (~5 MB)"] F["Fully Realized, Living Creature"] end A --> B A --> C A --> D A --> E B --> F C --> F D --> F E --> F ```

The game's texturing algorithms analyze the topological contours of the generated 3D mesh, automatically identifying features like the creature's spine or belly and painting them appropriately. The resulting creature can then be steered directly by the player or handed over to NPC artificial intelligence routines when imported into another person's game world.

Wright frames the entire procedural philosophy around the concept of creative amplification: - **Data Footprint**: The raw player-authored skeleton and part selections compress down to roughly 1 kilobyte of data. - **Runtime Expansion**: When the engine evaluates the procedural textures, polygon meshes, and kinematic animation rigs, the creature expands to the equivalent of a 5-megabyte asset. - **Compression Ratio**: This workflow achieves roughly a 5,000-to-1 data compression ratio. - **Practical Advantages**: This extreme data efficiency gives individual players extraordinary expressive leverage while allowing massive ecosystems of unique creatures to be stored in minuscule local databases, making deep gameplay possible even without an active internet connection.

[28:39](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1719s) Procedural Generation, Genotype Compression, and Content Pollination

Will Wright explains how the underlying architecture of *Spore* enables an unprecedented pipeline for content distribution. If an offline player installs the game, the installation CD comes preloaded with tens of thousands of procedural assets in an internal database. Because assets are represented not as heavy polygonal meshes or massive textures, but as compressed instructional recipes, they have almost zero friction when transmitted across networks, whether hosted on a central server or shared peer-to-peer. Wright points out that the creative handles given to the player are identical to the procedural parameters an automated system uses: the software's generative engine can turn the exact same algorithmic knobs that a human user adjusts when designing a character.

This design mirrors biological developmental systems. Just as in *The Sims*, where tweaking facial sliders triggers a procedural pipeline that compiles a final model, *Spore* treats user-defined slider values and bone configurations as a compact genetic code that undergoes embryological development to yield an interactive organism. This compact representation is the linchpin of what Wright calls "pollination." Whenever a player crafts a creature, that compact genotype is sent up to Maxis servers, where it is classified based on dietary habits, size, behavior, and life cycle. When a local game client requires rival species to populate an environment, it queries the central database through parametric matching, downloading suitable player-made creations to assemble a dynamic, balanced ecosystem on the fly. As Wright puts it: “And this is kind of basically mapping to the idea of a genotype going through embryology to an initial phenotype and that compression is vital for this whole thing.”

```mermaid flowchart LR accTitle: Spore Content Pollination Pipeline accDescr: Diagram illustrating how player creations become compressed genotypes, are categorized on a server, and are pulled dynamically into other players' local game worlds. subgraph Authoring["Client Authoring"] A["Player / Procedural Knobs"] --> B["Compact Genotype"] end subgraph Server["Central Processing"] B -->|"Low-friction upload"| C["Server Categorization\n(Diet, Size, Behavior)"] end subgraph WorldGen["Ecosystem Instantiation"] C -->|"Parametric query"| D["Embryology & Rigging"] D --> E["Phenotype in Local Ecosystem"] end ```

- Data transfer overhead remains minimal because the game transfers generative instructions rather than raw 3D geometry or baked texture maps. - Categorization algorithms track functional attributes—such as carnivore versus herbivore status, territorial traits, and scale—to ensure downloaded fauna naturally fit local ecological niches.

---

[30:08](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1808s) The Brain Palette and the Tribal Phase: Monkeys with Guns and Procedural Dance

Returning to the Creature Editor, Wright points to the final palette at the bottom of the user interface: the brain palette. Throughout the survival phase, players must resist the urge to spend all their earned points exclusively on immediate survival upgrades like bigger weapons, stealthy camouflage, or increased running speed. Instead, disciplined players divert resources into their brain capacity, treating it like a long-term college savings fund. Once the player invests enough capital to unlock the top tier of the brain ladder, their species crosses the threshold into sentience, triggering an abrupt mechanical genre shift. The game mutates from a first-person survival-eater into a real-time strategy (RTS) tribe simulator where the player manages a whole collective herd.

Wright references a recurring design landmark he introduced in previous lectures: his enduring creative obsession with "monkeys with guns." In the tribal phase, physical evolutionary morphology is locked in, and progression shifts toward cultural and material tooling. Wright purchases a spear rack, prompting excited, procedural vocalizations and animations as the pack inspects the weapon rack. To maintain order among his volatile creatures, he adds a campfire and a drum. Once the drum is placed, the creatures gather around and begin performing a dynamic dance routine whose animations are calculated entirely on the fly based on the creature's limb structure and joint topology, an effect that draws spontaneous applause from the conference audience. Play at this tier echoes classic god games like *Populous*: players assemble hunting expeditions, manage food stores, compete against rival tribes, and upgrade the central tribal hut to grow their population and unlock advanced tools. Wright emphasizes his design touchstone: “One of the landmarks is I've always been fascinated with the idea of monkeys with guns, and I don't know why, you know, but just what happens when you give monkeys guns, you know, it's just there's something cool about that.”

- Advancing the brain category permanently shifts the core game loop from individual creature survival to squad-based management. - Depending on player guidance and social reinforcement, a tribe can develop an emotional, expressive temperament or a cold, calculated, and logical culture. - Procedural verbs dictate creature interactions, enabling unique, morphology-specific animations for social activities, weapon handling, and musical performance.

---

[33:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1984s) The City Phase: SimCity Dynamics, Modular Architecture, and Cultural Aesthetics

Upgrading the central tribal hut to its final tier catapults the civilization into the City phase, which plays out as a streamlined, accessible incarnation of *SimCity*. Here, players zone roads, construct buildings, manage municipal services, and attend to the emotional happiness of their citizens. Every structure in the settlement is subject to complete authorial customization. When purchasing municipal structures, players can browse an integrated catalog pre-populated with creations downloaded from the community. An internal recommendation engine monitors the player's aesthetic choices and curates the storefront with assets that match their emerging stylistic profile.

For those who want to design from scratch, the building editor repurposes the morph-block interaction model seen in the creature creator, adapting it into a modular system reminiscent of digital Lego bricks. Players drag basic geometric primitives, intersecting or stacking them, while manipulating intuitive handles to reshape and extrude forms. Wright notes that this interface capitalizes on Maxis's discoveries during the development of *The Sims* build mode: giving users simple spatial operators and instant visual feedback creates an intrinsically rewarding toy. To demonstrate the aesthetic range of the toolset, Wright pans across the landscape to an adjacent rival city. While his own settlement features whimsical, organic silhouettes reminiscent of Dr. Seuss, the neighboring city boasts harsh, industrial, metallic architecture reminiscent of *Star Wars* or H.R. Giger. Wright emphasizes the player's complete artistic ownership over their community: “And basically I'm trying to keep them happy in their little city, you know, and again I have complete creative control over everything you see here, you know, so I am, you know, the player is the author of this little world.”

- Building editing relies on simplified CSG-like interactions (stacking and intersecting) anchored by contextual manipulation handles on platonic solids. - Stylistic range in the procedural building system spans an aesthetic spectrum from soft, stylized Pixar forms to hyper-detailed, industrial Giger architectures. - Citizen behavior mirrors structural quality and social stability, scaling up from simple interpersonal dynamics to crowd-level happiness indices.

---

[36:20](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2180s) Vehicles, City Specialization, and the Transition to Civilization

As municipal infrastructure expands, cities gain the capacity to manufacture vehicles, unlocking the primary vector for inter-city engagement. While tribal interaction relied on direct physical raids, trades, or cultural mingling, cities operate on a grander geopolitical scale. Players can subjugate or assimilate neighboring municipalities via three distinct strategies: military dominance, economic leverage, or cultural conversion. As an aggressive, industrialized neighboring city advances hostile units across the border, Wright enters the vehicle editor to craft a countermeasure.

```mermaid flowchart TD accTitle: City Interaction Vectors accDescr: Three distinct interaction vectors available to cities in the Spore City phase: Military, Economic, and Cultural. A["Player City Strategy"] --> B["Military\n(Direct conquest & armed force)"] A --> C["Economic\n(Trade dominance & financial buyout)"] A --> D["Cultural\n(Ideological influence & assimilation)"] B --> E["Rival City Assimilation"] C --> E D --> E ```

Like the architectural toolset, the vehicle editor combines modular styling with functional parts: attaching propellers enables flight, while specific chassis attachments grant aquatic or terrestrial capabilities. For demonstration purposes, Wright bypasses deep engineering and buys a goofy, Nazca-style flying contraption from the catalog. Upon deploying, the erratic craft flies aimlessly, prompting Wright to joke that his peaceful citizens are cowards who refuse to fly toward the battlefront. Despite the comic skirmish, Wright clarifies that brute combat is merely one optional path among more diplomatic and cultural alternatives. Panning the camera out past the immediate warzone, the local landscape rounds into a curved planetary horizon. This visual expansion marks the transition into the Civilization phase, where gameplay broadens into a global strategy experience modeled after games like *Civilization* and *Risk*. Wright outlines the tri-fold diplomatic framework: “At the city level we basically have kind of three ways to capture or ally with another city: they're basically military, economic, or cultural, depending on kind of the cultural flavor of my people.”

- Functional vehicle attachments directly dictate operational mobility across land, sea, and airspace. - Inter-city diplomacy is non-binary, allowing players to conquer via cultural proselytizing or corporate buyout rather than armed conquest alone. - Zooming out transitions the camera from a localized tactical perspective into a spherical planetary projection, shifting the scale to global geopolitics.

[38:10](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2290s) City Scale Transition, Micromanagement, and Database Pollination

Will Wright demonstrates how the game manages scale and micromanagement as the camera moves between strategic and tactical views. When zooming in close to a city, the buildings snap into realistic proportions; if the player wants to modify structures or continue designing their own version of Paris while under enemy attack, they can click individual buildings and open them directly in the editor. Conversely, pulling the camera back causes the city to exaggerate its proportions for both aesthetic and functional reasons, condensing down into an icon that represents a single interactive button with a streamlined menu. Wright explains that players will progress through this Civilization-style phase city by city and continent by continent until they have conquered and customized their entire home planet.

```mermaid flowchart LR accTitle: Spore Content Recommendation Loop accDescr: Diagram illustrating how player purchases feed back into database sorting to recommend content. subgraph Client ["Player Game Client"] Playstyle["Player Play Style"] BuyMode["Populated Buy Mode"] Purchase["Player Buys / Uses Content"] end subgraph Server ["Master Database"] Parametric["Parametric Mapping Engine"] MasterDB["Content Database Pool"] Sorting["Top-End Sorting Algorithm"] end

Playstyle --> Parametric MasterDB --> Parametric Parametric --> BuyMode BuyMode --> Purchase Purchase -->|"Positive Feedback"| Sorting Sorting -->|"Continuously Re-ranks"| MasterDB ```

Wright breaks down the procedural content sharing running underneath this stage, contrasting it with earlier phases. Rather than simple random sharing, the system draws from a master database of player creations and uses parametric mapping tailored to the user’s individual play style to populate their in-game "buy mode." When a player selects and purchases an asset created by someone else, that action serves as positive feedback for the central database. The server continuously resorts and re-ranks assets at the top end based on broad player adoption, curating the highest-quality community creations automatically.

[39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) Genre Mixing, Design Philosophy, and Player as Author

Addressing the overall game structure, Wright compares Spore to *WarioWare, Inc.*, describing it as an unapologetic homage to his favorite strategy games. He notes that conventional industry wisdom warns developers never to mix disparate genres:

> “And I know it's kind of like one of the rules of thumb in the game is don't mix genres, and so I've always wanted to break that rule really badly.”

He details the classic influences underpinning each stage: - **Tide Pool Phase:** Draws heavily from *Pac-Man*. - **Creature Evolution Phase:** Mirrors the dynamic loop of *Diablo*, where the player ventures out into a hostile world to fight monsters, returns to the safety of an editor (acting as the village) to upgrade their creature's "armor and swords" across generations, and ventures back out to battle harder threats. - **Tribal Phase:** Built as an explicit homage to Peter Molyneux's *Populous*. - **City Phase:** Distills the essence of *SimCity*. - **Civilization Phase:** Borrows the empire-building spirit of *Civilization*.

Wright clarifies that he wanted each phase to offer a distilled, lighter experience—around 20 percent of the best mechanics of *SimCity* or *Civilization*. As an aging hardcore gamer, he finds himself playing handhelds like the Game Boy because he no longer has dozens of hours to invest in massive systems. Wright anticipates that roughly 40 percent of the Spore experience will be purely aesthetic creative expression, supported by simple, accessible systems that anyone can pick up. Crucially, the editors attached to each level reshape the player's agency:

> “And really, you know, I did not want to put the player in the role of, you know, Luke Skywalker or Frodo Baggins; instead I wanted to put the player in the role of George Lucas or J. R. R. Tolkien or Dr. Seuss, or whoever, you know? I wanted the player to feel like, you know, they had built an entire world, you know, going through these different levels.”

[41:49](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2509s) Emergent Storytelling and the Robert Louis Stevenson Model

Wright pauses to reflect on narrative design, arguing that narrative should be treated as an emergent result of interactive systems rather than a rigid baseline:

> “More recently I've been thinking of story as really a side effect or a property of interesting experiences, not a prerequisite.”

He points out that games like *Grand Theft Auto* often provide incredible personal player stories despite the player ignoring the scripted plot; because the events happen directly to the player, they possess a strong sense of personal ownership. In Wright's view, stories serve as filtering processes that emphasize causality. Unlike cinema, where a director can cleanly sequence cause and effect—such as showing precisely why a character gets angry, kicks an object, and breaks their foot—games rely on the player's personal experience of the events to provide that logic. Stories also function as tools for portability, compression, and abstraction, converting concrete gameplay events into mental schemas that apply to other contexts. Wright recalls Robert Louis Stevenson’s process writing *Treasure Island*, noting that Stevenson began simply by drawing an island map and letting his eyes wander across the terrain, allowing the narrative to emerge naturally from the geography.

[43:22](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2602s) The UFO Swiss Army Knife, Abduction, and Interplanetary Scale

Returning to the live game, Wright looks over his stylized "clock-like toy world" and inspects the local city, noting a fresh crater where an attacked building once stood. He enters the vehicle editor to highlight its final unsealed category: after unlocking ground, water, air, and submersible craft, players eventually earn the UFO. Wright selects a stock flying saucer, describing it as the ultimate Swiss Army knife for the late game, granting unmatched speed to traverse the globe instantly.

He shares a childhood memory that inspired the vehicle's mechanics: watching *Mutual of Omaha's Wild Kingdom*, hosted by Marlin Perkins. Wright recounts that no matter the crisis or endangered animal being studied, the climax almost inevitably involved leaning out of a helicopter with a high-powered tranquilizer gun to dart a creature. Channels of that show inspired Spore's abduction beam, which Wright uses to capture wildlife. Scouting the planet's surface, he spots the hostile creature species that had bullied his tribe earlier in the game, declaring it time for payback. After a brief chase, he lines up the beam, snatches a specimen directly off the terrain, and stores it in his cargo hold.

Wright then pulls the camera straight up, flying higher and higher as the planet shrinks, pulling the audience past the atmosphere until the entire solar system comes into view, earning enthusiastic applause from the room. Wright points out the system's comets, the central sun, hot planets orbiting close in, the home world, and frozen outer bodies. He zooms the camera toward a distant, barren moon with radically different geology, explaining that every planet functions as a blank-slate sandbox for player experimentation and colonization. Seeking to bootstrap the frozen moon with life, he descends to its desolate surface and drops the creature he just abducted from home, only for it to instantly perish in the harsh environment. As Wright asks the audience why the creature died, he calls over to the audio booth to turn up the stage machine's volume for the next demonstration.

[47:43](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2863s) Terraforming and Colonization Mechanics

Addressing a world lacking an atmosphere, Will Wright explains that the overarching objective is to transform barren planets into livable habitats. In conceptualizing the terraforming toolset, he drew creative inspiration from the classic educational painting program *Kid Pix*. Wright appreciated that *Kid Pix* gave users wacky, expressive tools that constantly risked slipping out of strict control, producing serendipitous accidents. As Will Wright puts it: “Basically what I loved about Kid Pix is you had all these cool weird wacky creative tools and they were always out of your control—I mean you always did these unexpected oops I didn't mean to do that type of thing—and I wanted the terraforming to kind of feel like that as well.”

```mermaid flowchart TD accTitle: Planetary Terraforming and Colonization Loop accDescr: Diagram illustrating the iterative loop between terraforming, establishing colonies, generating revenue, and purchasing advanced UFO upgrades.

subgraph Terraforming["Atmospheric Engineering"] A["Barren Planet (No Atmosphere)"] --> B["Volcano Tool / Gas Emission"] B --> C["Atmosphere Densifies & Biosphere Score Rises"] C --> D["Plant Seeding & Biosphere Maturation"] end

subgraph Colonization["Colony Infrastructure"] A -.->|"Early Settlement"| E["Enclosed Bubble Colony"] D --> F["Open-Air Elaborate Cities"] end

subgraph Economy["UFO Progression"] E --> G["Earn Revenue / Resources"] F --> G G --> H["Purchase Advanced Tools & Genesis Device"] H -->|"Accelerated Seeding"| D G --> I["Unlock Interstellar Drive"] end ```

To demonstrate, Wright deploys a volcano tool directly onto the planetary crust, simultaneously altering the elevation of the terrain and venting gases into the sky. Over time, these thermal vents generate a dense atmosphere. While waiting for the atmospheric pressure to stabilize, players can immediately begin founding settlements. Early colonies on airless rocks appear beneath sealed biodomes—a visual motif shared with deep underwater cities on oceanic home planets. Once an atmosphere is fully formed, domes are discarded in favor of sprawling open-air colonies, which generate revenue to finance further UFO equipment upgrades. To bypass the natural timeline during the keynote demo, Wright triggers an endgame superweapon called the Genesis Device, instantly clearing storm clouds, greening the landscape, and boosting the planet's biosphere score so harvested flora and fauna can take root.

- **Kid Pix Philosophy**: Emphasizes unpredictable, tactile, and expressive creative tools over rigid landscape editors. - **Atmospheric Mechanics**: Volcanic activity pumps out gases to alter density, cloud cover, and surface color. - **Bubble Architecture**: Applied interchangeably to airless extraterrestrial colonies and domestic oceanic underwater settlements. - **Economic Loop**: Settled colonies produce funds that allow players to purchase higher-tier ship capabilities and terraforming gadgets. - **Genesis Device**: A late-game, high-cost item that instantly bootstraps planetary atmospheres and initiates rapid biosphere growth.

---

[51:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3060s) The Stellar Zoo and Asynchronous First Contact

After accumulating sufficient capital from provincial colonies, players can purchase the interstellar drive to break out of their solar system and enter the galactic view. Wright terms this macro-environment the "stellar zoo." He expresses deep frustration with how traditional astronomy curricula flatten outer space into two-dimensional star charts, obscuring the three-dimensional grandeur revealed by deep-space photography. As Will Wright explains: “One thing that's always really annoyed me about the way astronomy is presented is it's presented as two-dimensional—if you look at star maps like that, I mean that's the way astronomy is taught—but it's yet it's this beautifully three-dimensional thing and you just are not given the sense of what that's like.” The galactic map incorporates authentic cosmological entities—such as black holes, planetary nebulae formed from supernova shells, star-forming stellar nurseries, and hazardous blue giant stars on the verge of detonation.

Navigating this interstellar terrain, the player uses a SETI radio scanner tool to mouse over stellar targets and listen for artificial transmissions. Picking up an alien broadcast, Wright pinpoints an inhabited solar system and descends upon an arid, sun-scorched planet populated by a civilization designed by an entirely different human player. Wright emphasizes that *Spore* does not run as a synchronized real-time multiplayer game; instead, worlds and species are pollinated asynchronously across the network. When visiting a downloaded world, the local AI assumes control of the foreign species, driving their behavior using the personality sliders chosen by the player who originally evolved them.

Wright approaches a large alien metropolis to initiate first contact, launching an interactive musical dialogue inspired by *Close Encounters of the Third Kind*. When the musical exchange fails and the hostile natives scramble anti-aircraft fighters in an arcade-style dogfight reminiscent of *Space Invaders*, Wright tests the UFO's offensive loadout. Finding basic energy beams underwhelming, he deploys an ultimate weapon that completely shatters the alien planet into an asteroid ring. Observing the catastrophic planetary destruction, Wright dryly jokes: “So there you go, our friends—and there goes my reputation as a non-violent game developer.”

- **3D Astronomy Sandbox**: Incorporates genuine astrophysical objects, including supernova shells, stellar nurseries, blue giants, and black holes. - **SETI Audio Scanning**: Players scan stellar coordinates with directional radio receivers to detect intelligent alien transmissions. - **Asynchronous Pollination**: World states, creatures, and structures seamlessly download in the background; offline AI pilots them based on original player sliders. - **First Contact Minigame**: A musical pitch-matching mechanic modeled on *Close Encounters of the Third Kind* determines diplomatic outcomes. - **Extinction-Level Escalation**: Ineffective dogfighting weapons give way to planetary demolition tools that collapse entire worlds into rings of debris.

---

[55:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3321s) Science Fiction Metagames and the T-Shaped Architecture

The space stage functions as a massive canvas designed around classic science fiction tropes. Wright lists several distinct metagames players can pursue: invasion battles inspired by *The War of the Worlds*; peacekeeping missions reminiscent of *The Day the Earth Stood Still*; monolithic uplifting of primitive creatures inspired by *2001: A Space Odyssey*; genetic abduction and hybridization; and deep galactic diplomacy. To make race relations distinct, Maxis mapped planetary personalities along behavioral dimensions inspired by archetypal *Star Trek* civilizations, ensuring that belligerent Klingon-like societies clash naturally with relentless, expansionist Borg analogues. Visually, Wright drew inspiration from 1930s and 1940s pulp sci-fi magazine covers, which celebrated the very combination of elements *Spore* systematizes: exotic vehicles, retro-futuristic architecture, bizarre aliens, and alien planetary vistas.

```mermaid flowchart TD accTitle: The T-Shaped Game Progression Model accDescr: Flowchart illustrating Spore's T-shaped architecture, showing how linear evolution stages serve as tutorials feeding into the expansive horizontal and vertical space sandbox.

subgraph Stem["Vertical Evolution (Tutorial / Goal-Oriented Progression)"] T1["Tide Pool / Cellular Stage"] --> T2["Creature Evolution Stage"] T2 --> T3["Tribal Stage"] T3 --> T4["City / Civilization Stage"] end

subgraph Crossbar["The Space Sandbox (Horizontal & Fractal Freedom)"] T4 --> S1["Interstellar Space Travel"] S1 <--> M1["Diplomacy (Star Trek Archetypes)"] S1 <--> M2["Abduction & Cross-Breeding"] S1 <--> M3["Uplift (Monolith Technology)"] S1 <--> M4["Conquest & Galactic War"] end

Crossbar -.->|"Fractal Deep-Dive Surfing"| Stem ```

At this macro level, content pollination expands from procedural distribution to overt browsing. Players can visit specific user-generated planets, abduct native wildlife, hijack signature vehicles, and transport specimens back home to seed custom planetary zoos. This loop completes what Wright describes as the overarching "T-shaped" structural design of *Spore*. Rather than following the conventional game industry paradigm—where open sandboxes serve as simple, low-stakes tutorials before funneling users into restrictive goal-oriented endgame campaigns—*Spore* deliberately turns the model upside down. As Will Wright points out: “We're using the goal-oriented game to basically train you to use all the editors and to teach you the simulation dynamics at every level, so by the time you get to the space game you can kind of fractally surf, you know, both horizontally, you know, from planet to planet, and vertically, you know, from like city to civ to tribal.”

- **Sci-Fi Archetypes**: Metagames incorporate themes of planetary uplifting, alien abduction, first contact, interspecies warfare, and diplomacy. - **Personality Dimensions**: Alien behavior is calibrated against recognizable science fiction archetypes, using the behavioral spectra of *Star Trek* races to balance relations between warmongers, traders, and diplomats. - **Pulp Aesthetic Heritage**: The visual range of vehicles, buildings, creatures, and landscapes directly channels the cover art of 1930s–1940s science fiction magazines. - **Active Pollination & Specimen Harvesting**: Players can beam foreign flora and fauna into their hold to stock customized planetary nature reserves. - **Inverted Sandbox Architecture**: Progression moves up a narrow stem of goal-directed evolution stages (Cell, Creature, Tribe, Civilization) that teach the editors, culminating in an expansive horizontal and vertical space sandbox.

[58:27](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3507s) The Genesis of the Concept: *Powers of Ten*, SETI, and the Drake Equation

Looking at the galactic overview, Will Wright explains that the entire civilized planet shown earlier occupied just a single microscopic square in an immense galaxy consisting of tens or hundreds of thousands of worlds, all open for players to visit, alter, and make up their own games within. The concept traces its origins back to Charles and Ray Eames' seminal short film *Powers of Ten*, of which Wright had long been an admirer. For years, however, he had struggled to conceptualize how to unify that dramatic progression across physical orders of magnitude into a coherent, functioning game structure.

The breakthrough arrived through Wright's fascination with astrobiology and an independent prototype he had started exploring around the Search for Extraterrestrial Intelligence (SETI). Examining Frank Drake’s famous formula for estimating the number of active, communicative extraterrestrial civilizations in the Milky Way, Wright realized that the individual variables within the Drake equation—from the rate of star formation and fraction of stars with planetary systems, to the emergence of life, intelligence, and communicative civilizations—mapped seamlessly onto distinct exponential powers of physical scale. By folding the SETI concept and the *Powers of Ten* framework into a single continuous design, the foundation for *Spore* was born.

In a wry aside about industry secrecy, Wright reveals that the game's earliest design document had actually been hiding in plain sight for an entire year. *Wired* magazine had approached him to design a single page for an upcoming issue focused on the theme of life, offering him total creative freedom. Wright simply drafted and submitted the foundational design schematic for the game without identifying what it was, and the magazine published it without anyone in the gaming community deciphering its true purpose.

```mermaid flowchart TD accTitle: Spore Conceptual Synthesis accDescr: Diagram illustrating how Powers of Ten and the Drake Equation merged into Spore's multi-scale architecture. subgraph Inspirations ["Inspirational Frameworks"] A["Eames' 'Powers of Ten'<br/>(Scales of the Universe)"] B["SETI & Astrobiology<br/>(The Drake Equation)"] end subgraph Realization ["Structural Alignment"] C["Equation terms map directly<br/>to exponential physical scales"] end subgraph Execution ["Spore Progression"] D["Microscopic Tidepool<br/>(Bacteria)"] --> E["Planetary Civilization<br/>(Culture & Tech)"] E --> F["Galactic God<br/>(Interstellar Empire)"] end A --> C B --> C C --> Execution ```

[59:47](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3587s) Compressing Life: The Architecture of Generative Content

To make such an intimidating scope understandable to the broader public, the team distilled the player's core journey into a clean, intuitive arc: guiding an organism from a single bacterium up into a galactic deity, effectively simulating the full past and speculative future of life. Wright jokes about the difficulty of condensing that ambition into typical publishing parlance, remarking, “It's hard to come with a really kind of catchy marketing line for that if anybody comes up with one let me know please.”

The mechanical linchpin binding this entire simulation together is algorithmic compression and procedural generation. Rather than forcing artists to hand-craft static assets, procedural compression provides immense creative leverage to the player, allowing anyone to fashion idiosyncratic creatures, buildings, and vehicles that the engine instantly animates and parses. That same structural leverage feeds back into generative systems that populate massive shared content libraries. Because the procedural data files represent compact mathematical instructions rather than raw polygonal meshes and textures, assets become extraordinarily lightweight and portable, allowing the client to continuously stream thousands of player-created species across an unbounded universe.

Wright emphasizes that reaching this stage required immense technological exploration: - Maxis built over 200 distinct experimental prototypes over the course of development. - Wright had actually displayed several of these isolated prototypes during earlier talks and lectures over the years without ever disclosing their ultimate application. - Iterative prototyping became what Wright describes as an organizational “religion” to validate gameplay systems that had never existed before.

[01:00:54](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3654s) The Bottleneck of Imagination and the GDC Farewell

Reflecting on his career developing unconventional titles, Wright observes that creators often assume their chief obstacle lies in external resistance—convincing skeptical executives, publishers, or focus groups that a strange title can actually sell. On this project, however, the real roadblock was internal. As Wright admits to the audience, “When I look back on this idea you know the biggest obstacle I truly had was making myself believe that we could build the game it was my own imagination was the biggest bottleneck.” Once he crossed that threshold of conviction and proved to himself that the mathematics and technology were viable, bringing his development staff and company leadership on board proved surprisingly straightforward.

He urges the developers in the room not to discard their most radical impulses simply because the production hurdles appear insurmountable on first glance. When designers hit upon an eccentric idea that seems too far outside conventional industry paradigms to possibly succeed, they should continually revisit it, as technological breakthroughs and shifts in perspective can eventually make the impossible practical.

Wright closes the presentation on a conspiratorial note, explaining that Electronic Arts had granted him strict, one-time permission to reveal the game exclusively to his peers in the room at GDC. He warns that he is forbidden from uttering another word about it publicly until the Electronic Entertainment Expo (E3) later that year, before stepping away from the podium to thunderous, sustained applause from the auditorium, playfully telling the audience, “You can keep clapping.”

## Notable quotes > [05:24](https://www.youtube.com/watch?v=ofA6YWVTURU&t=324s) I contend that player stories will always be more powerful than scripted stories that we try to tell the players. — Will Wright

> [39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) I did not want to put the player in the role of, you know, Luke Skywalker or Frodo Baggins; instead I wanted to put the player in the role of George Lucas or J. R. R. Tolkien or Dr. Seuss, or whoever, you know? — Will Wright

> [41:49](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2509s) More recently I've been thinking of story as really a side effect or a property of interesting experiences, not a prerequisite. — Will Wright

## Mentioned Will Wright, Maxis, Electronic Arts, Game Developers Conference, Spore, The Sims, The Sims 2, SimCity, Grand Theft Auto: San Andreas, Ocean Quigley, Kid Pix, Civilization, Populous, Pac-Man, Diablo, Myst, Cyan, George Lucas, J. R. R. Tolkien, Dr. Seuss, H.R. Giger, Pixar, SETI, Frank Drake, Powers of Ten, Charles and Ray Eames, Robert Louis Stevenson

## Video description Original Video Description: Will Wright talking at the 2005 Game Developer's Conference about 'Spore', which looks like it could possibly be the best video game ever.

https://www.gdcvault.com/play/1019981/The-Future-of-Content-(English

Source Videos: https://archive.org/details/2005_GDC_-_Will_Wright_SPORE_Reveal_-_Camera_Angle_of_Wright_Only_-_2005-03-11_A https://archive.org/details/2005-gdc-will-wright-spore-demonstration-audience-reaction https://archive.org/details/2005-will-wright-keynote-spore-screen-only-2005-03-11

I tried to fix up this video as best as I could. Unfortunately the game sections were recorded off of a projector screen so it's pretty blown out.

Original Videos (Dead Links): http://video.google.com/videoplay?docid=8372603330420559198 http://video.google.com/videoplay?docid=-775656482294094003

[Source](https://www.youtube.com/watch?v=ofA6YWVTURU)

中文简报

威尔·赖特公布《孢子》以应对次世代暴涨的内容预算 程序化生成与玩家自动化创作将轻量级玩家指令转化为功能完备的世界,无需大幅增加工作室人员规模,从而化解游戏行业次世代预算危机。 针对业界对次世代资产制作成本急剧膨胀的担忧,威尔·赖特提出程序化生成与玩家能动性提供了一条可扩展的发展路径,让消费者也能成为世界的创作者。在展示《孢子》的早期Alpha原型时,他呈现了横跨天文尺度的无缝推进——从2D细胞潮池到陆生生物进化、市政建筑,再到完整的星际探索。赖特详细说明了程序化反向运动学、行为合成以及自动化“盒中艺术家”管线如何实现5000比1的压缩比,仅通过轻量级数学指令集(约1 KB)即可异步充实一个共享的银河生态系统。 威尔·赖特在GDC 2005上公开了《孢子》的原型,正面挑战了当时业界的共识,即次世代高清游戏将为开发者带来不可持续的资产制作危机。 (0:00) 《孢子》通过程序化生成避开了庞大的存储需求,将复杂的玩家造物存储为约1 KB的参数集,并在运行时解构成约5 MB的资产——实现了5000比1的压缩比。 (26:35) 该游戏基于异步多元宇宙模型运行,玩家创作的生物、建筑和载具会根据参数化查询从中央服务器动态下载,以充实其他玩家的生态系统。 (14:35) 《孢子》结合了查尔斯和雷·伊姆斯纪录片《十的乘方》中的指数级宏观推进与搜寻地外文明计划(SETI)的德雷克方程,引领玩家从单细胞生命一路演进至银河文明。 (58:27)

Reading summary

At GDC 2005, game designer Will Wright unveiled *Spore* to argue that the video game industry's looming content crisis—driven by the escalating labor costs of high-definition hardware—can be solved by replacing pre-baked assets with procedural generation and player-authored content. The Production Trap and Player Agency (2:04) Wright argues that the industry's transition to next-generation hardware creates an economic bottleneck: studio teams expand disproportionately to produce handcrafted art assets, but standard utility theory dictates that doubling content volume yields diminishing returns for players. In contrast, player communities generate vast volumes of content when provided tools; Wright notes that *The Sims 2* official exchange quickly accumulated 17,000 characters and 10,000 houses, while a fan site logged its 100,000th custom object. He contends that evaluating games solely on a two-dimensional grid of content quantity versus visual fidelity overlooks a vital third dimension: player ownership and agency, which produces far deeper engagement and personal storytelling than pre-scripted cinematic sequences. Procedural Synthesis and 'Artist in a Box' (6:41) Tracing the history of game programming, Wright notes that early Apple II development split memory evenly between code and data, but the arrival of the 650-megabyte CD-ROM led the industry to prioritize massive pre-rendered data files over algorithmic efficiency. Inspired by the demoscene's ability to fit complex audiovisual routines into 64-kilobyte executables, Wright's team adopted procedural techniques to build an automated pipeline dubbed 'artist in a box.' By automating mesh generation, UV unwrapping, texturing, and rigging, the engine amplifies minimal player inputs into complete assets: a player's skeletal sketch of roughly 1 kilobyte dynamically expands into a fully rigged, textured, and physics-driven in-game entity. Genotype Compression and Content Pollination (28:39) Rather than storing and transmitting heavy polygonal models, *Spore* compresses player-authored assets into compact instructional recipes, or genotypes, mirroring biological embryology. This compression enables asynchronous multiplayer ecosystems through what Wright terms 'content pollination.' When an individual plays the game, their local client automatically uploads their designs and downloads parametric matches authored by other players to populate balanced ecological niches, dynamic architectural styles, and rival planetary empires. Feedback mechanisms track player interactions to continuously sort and recommend high-quality community assets without requiring manual web downloads or peer-to-peer file swaps. Iterative Scale from Bacteria to Galactic Diety (39:48) Structurally, *Spore* synthesizes Charles and Ray Eames' short film *Powers of Ten* with the variables of Frank Drake's astrobiological equation, linking each physical scale to classic game mechanics. The experience advances along a T-shaped design progression: linear vertical phases function as accessible tutorials—drawing on *Pac-Man* for cellular survival, *Diablo* for creature evolution, *Populous* for tribal leadership, *SimCity* for municipal planning, and *Civilization* for global geopolitics—before opening into an unbounded horizontal space sandbox. In the space phase, players pilot terraforming and abduction-capable UFOs across a galactic 'stellar zoo' populated by alien species that act according to personality profiles established by other players. Wright concluded that the primary hurdle in developing *Spore* was overcoming his own doubts about whether the underlying procedural mathematics and systems could be implemented. Validating the game's uncharted mechanics required Maxis to build over 200 isolated experimental prototypes, leading Wright to urge fellow developers not to abandon radical concepts simply because production hurdles appear insurmountable under current industry conventions.

精炼阅读摘要

在2005年游戏开发者大会(GDC)上,Will Wright针对次世代硬件所引发的昂贵劳动力危机,提出以程序化生成与玩家自建内容打破工业瓶颈,并通过其新作《Spore》展示了从小尺度单细胞到星际文明的生成式游戏构想。 次世代内容危机与玩家所有权维度 (2:04) Wright指出,游戏业当时普遍担忧次世代硬件带来的高昂人力成本将扼杀中小厂商与创新。依据效用理论,单纯堆砌双倍的手工制作资产需投入翻倍精力,却难以换得成比例的情感回报。他以《The Sims 2》官方平台上玩家自发上传的1.7万名角色与1万座房屋为例,证明社群的创作产能远超工作室内部手工生产。行业过去仅将内容价值简单归结为“数量×质量”的二维模型,Wright提出必须加入第三个维度——“玩家的所有权与主体性”。通过赋予玩家设计权,让系统记录并重组其个人叙事,不仅能规避不合理制作成本,还能带来远比传统预设脚本更深刻的情感投入。 从黑客演示到算法复兴与盒中艺术家 (6:41) 回顾开发历史,Wright提及Apple II时代由于存储空间极其有限,程序与数据各占约50%的内存;而650MB容量的CD-ROM出现后,这种平衡被彻底打破,海量手工预渲染数据压制了算法设计。以北欧北极圈一带的Demoscene黑客群体为灵感,他们用不到64KB的二进制文件生成庞大的动态音画,Wright团队以此确立了“盒中艺术家”构想。该系统旨在将网格生成、拓扑分析、程序化贴图绑定以及基于反向运动学的骨骼动画全流程自动化。在《Spore》中,玩家简单的结构编辑输入仅产生约1KB的基因型数据指令,经由引擎在运行时将其展开为体积达数兆字节的动态完整生物。 异步内容授粉机制与微观至宏观的玩法演进 (14:35) 《Spore》的核心机制依赖其异步数据分发系统。当单机玩家进入世界时,客户端并不会从中心服务器同步实时联机状态,而是将其他玩家上传并经过行为与尺寸分类的紧凑“基因型”资产作为授粉源,拉取至本地以填补生态位或城市商店。游戏的进程跨越多个经典玩法阶段:在单细胞与潮池阶段,玩家以类《Pac-Man》的形式吸收养分并动态组装进攻或游动器官;在陆生生物阶段,动作模块实时结合运动学生成如拖拽猎物等自发行为;进入部落与城市阶段后,游戏分别汲取了《Populous》与《SimCity》的策略和乐高积木式建筑编辑逻辑,直至最后以类《Civilization》的模型统一整个行星。 太空沙盒与物理尺度融合的概念源流 在游戏的终局太空阶段,飞行载具作为多功能平台,允许玩家通过类似经典绘画软件《Kid Pix》风格的非严密控制工具进行行星地球化改造,包括释放火山气体制造大气层与建立隔离生态圈。该宏观宇宙结合了天文学真实结构与经典科幻命题,玩家可在数以万计的星系间展开外交通信、生物绑架与文明提升。Wright将这种从微观至宏观的T型架构归功于Charles和Ray Eames制作的纪录片《Powers of Ten》以及探索地外智慧生命的德雷克公式,他将公式中从恒星形成到智能文明诞生的各项指数级变量,一一映射到相对应的连续物理尺度系统之中。 Wright强调,尽管团队为验证跨维度玩法构建了200多个原型,但颠覆性设计最大的瓶颈往往并非出版商阻力或硬件限制,而是创作者自身想象力的自我怀疑;《Spore》的技术可行性证明了算法模型有能力跨越传统工业化内容管线的沉重制约。

Detailed notes — zh

## Background 该视频由 YouTube 频道 Zeusmonomon 于 2020 年 8 月 27 日上传发布,是对 2005 年游戏开发者大会(GDC 2005)现场演讲的重置修复版本。视频简介指出,传奇游戏设计师威尔·赖特(Will Wright)在本次大会上发表了题为《内容的未来》(The Future of Content)的演讲,首次向业界揭晓了划时代大作《孢子》(Spore)的技术与设计理念。 ## TL;DR 威尔·赖特在演讲中反思了从《模拟人生》开发中汲取的经验,指出次时代游戏制作中美术资产膨胀与边际效用递减的危机,并提出以算法驱动的过程化生成(Procedural Generation)作为破局解法。通过向北欧 Demo Scene 的极限压缩技术取经,团队打造了一套将 1 KB 玩家输入放大为 5 MB 动态资产的“盒中艺术家”管线,实现了生物动画、纹理与生态的自适应演化。演示中展示了《孢子》从单细胞、三足陆生生物、部落与城市文明,一路无缝缩放至星际殖民与外星遭遇的宏大跨度,倡导以轻量化数据授粉与玩家绝对归属感为核心的全新游戏范式。 ## Atmosphere & style 现场弥漫着技术开拓者所特有的极客幽默与惊叹氛围。威尔·赖特语速飞快、风趣自嘲,毫无商业宣发的做作,像一位充满童心的老顽童般向同行拆解自制原型中的各种奇思妙想。每当演示镜头从微观细胞突破常规、层层向外拉升至浩瀚银河与行星球体时,台下开发者同行都会爆发出一阵阵真实的狂喜与雷鸣般的掌声,整场演讲兼具深度的工程洞察与令人屏息的浪漫愿景。 ## Bulletin - [00:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=0s) Will Wright 在 GDC 2005 上首度公开代号为《孢子》(Spore)的原型演示,提出以过程化生成(Procedural Generation)化解次时代游戏人工内容堆砌与成本暴涨的危机。 - [26:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1595s) 《孢子》利用“盒中艺术家”算法实现高达 5000:1 的数据放大比:玩家保存的轻量化参数仅约 1 KB,引擎在运行时可动态解压展开为约 5 MB 包含骨骼、材质网格与 AI 行为的完整游戏实体。 - [14:26](https://www.youtube.com/watch?v=ofA6YWVTURU&t=866s) 游戏通过中央服务器的“授粉”(Pollination)机制实现异步生态系统分发,根据单机生态位缺口自动拉取全球其他玩家创造的生物与建筑,无需同步联机即可编织出动态平衡的宇宙。 - [55:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3321s) 系统采用倒置的“T 型”游戏架构:细胞、生物、部落和文明等纵向目标阶段实为教程,旨在引导玩家掌握工具与模拟规则,最终冲顶解锁横向无限延伸的太空沙盒与科幻元游戏。 ## Key takeaways - [02:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=124s) 传统内容生产存在边际效用递减规律,单纯堆叠人工资产在经济学上不可持续,而玩家自制内容(UGC)在体量与创造力上展现出碾压官方人工制作的潜力。 - [02:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=124s) 游戏感知的价值维度不仅由“数量”与“质量”构成,关键的第三维度是“归属感”(Ownership);玩家亲自创造角色的自发故事远比开发者预设的脚本剧情更具感染力。 - [06:19](https://www.youtube.com/watch?v=ofA6YWVTURU&t=379s) CD-ROM 等大容量介质的普及打破了 Apple II 时代代码与数据 1:1 的平衡,使工业界滑向数据至上与预制资产堆叠,敲响了传统纯算法设计的丧钟。 - [08:30](https://www.youtube.com/watch?v=ofA6YWVTURU&t=510s) 北欧 Demo Scene 群体仅凭 64 KB 的极限代码实时演算视听奇观的极客文化,为《孢子》全盘采用纯算法与过程化生成提供了核心技术灵感。 - [15:52](https://www.youtube.com/watch?v=ofA6YWVTURU&t=952s) 《孢子》的生物运动摒弃了预制动画资产,依托“程序化动词”与受力拓扑分析,无论玩家组装出何种怪异形态,算法均能实时演算出与之匹配的自适应步态与动作融合。 - [28:39](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1719s) 游戏机制映射了生物学过程:数据参数如同“基因型”(Genotype),经由胚胎发育般的算法实时解压生成游戏世界内的“表现型”(Phenotype),大幅降低了网络分发与存储阻力。 - [30:03](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1803s) 智慧觉醒推动玩法类型范式转移:在“脑力面板”沉淀投资突破阈值后,游戏从第一人称视角的生存吞噬无缝跃迁至类似《上帝也疯狂》的部落即时战略模式。 - [33:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1984s) 建筑与载具编辑器引入乐高式的正多面体自由塑形,组件兼具物理功能属性,支撑城邦间军事征服、经济渗透与文化同化等多元演进路线。 - [39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) 游戏有意混搭各流派精髓(《吃豆人》《暗黑破坏神》《模拟城市》《文明》等),力求提炼各经典最核心的 20% 轻量化策略,将 40% 的体验留给玩家的美学创意。 - [41:49](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2509s) 故事不应作为游戏的前提预设,而是交互因果与体验自然涌现的副产物;玩家的角色定位应从“卢克·天行者”转变为创世者“乔治·卢卡斯”。 - [43:29](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2609s) UFO 是后期集大成的“瑞士军刀”载具,融合了绑架光束、地形改造(Kid Pix 风格)与“创世纪装置”,支持玩家改造行星大气并建立穹顶殖民地。 - [51:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3060s) 三维银河空间跳脱传统教科书的平面星图局限,融入黑洞、超新星遗迹及危险蓝巨星,并通过 SETI 雷达监听外星通讯以触发跨玩家异步遭遇战。 - [58:27](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3507s) 《孢子》的宏大跨度源于将查尔斯和雷·伊姆斯夫妇纪录片《十的次方》的尺度跳跃,与天文学德雷克方程的各级参数一一映射,将生命演进与星际探索贯通为一体。 - [59:47](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3587s) 为了攻克前所未有的生成式体系,团队前后打造了 200 多个独立原型,证明了在技术攻坚中打破自身想象力瓶颈、坚守边缘创意的关键价值。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=0s) 开场介绍与演讲题目的“谎言” 在热烈的掌声与欢呼声中,主持人宣布演讲开始,并向观众隆重介绍了 Will Wright。Will Wright 走上讲台,打趣地确认麦克风的声音是否如同“上帝之声”(voice of god),随后配合主办方的要求宣布了关掉手机、禁止拍照以及填写反馈表等现场规则。他首先介绍了在台上一同协助演示的联合设计师 Chris,并提到两人已经并肩合作了大约五年。 面对台下的国际观众,Will Wright 特别向同传翻译员表达了歉意,坦承自己讲话语速极快,总想在短短一小时内塞入海量信息,并开玩笑说:“如果你们跟不上,随便编点什么都行,说不定比我讲得还要精彩。”谈到本次演讲的标题与简介时,他笑着承认那完全是个“绝望的含糊之词”。往年 GDC 邀请他演讲时,他通常在过了一周后才会想出一个大概的方向,然后交出一个通用的泛泛标题。然而今年却截然不同:“今年当他们问我打算讲什么时,我直接彻底撒了谎。”他直言,本次分享并不是去泛谈所谓“内容的未来”,而是探讨一种“可能的内容未来”。在本届 GDC 上,整个行业弥漫着一种悲观论调:次时代大作的研发成本正呈天文数字上涨,开发者必须堆砌海量内容,这可能将大多数中小型团队彻底挤出市场。Will Wright 希望提出一种完全不同的替代方案,探索更具创造力的方式来应对内容膨胀危机。 ### [02:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=124s) 从《模拟人生》学到的内容经济学与玩家归属感 ```mermaid flowchart LR accTitle: 玩家价值维度模型 accDescr: 展示传统内容价值模型与加入玩家归属感维度后的对比 subgraph Traditional["传统二维评估模型"] Q1["内容数量 (Quantity)"] --> A1["感知价值面积 (Player Value)"] Q2["内容质量 (Quality)"] --> A1 end subgraph Expanded["三维玩家驱动模型"] A1 -.->|"引入全新维度"| V3D["高阶感知价值空间"] O["归属感与玩家创作 (Ownership)"] ==> V3D end ``` Will Wright 认为,本次演讲更准确的标题或许该是《我从〈模拟人生〉中学到的内容二三事》(What I Learned About Content from The Sims)。通过展示制作《模拟人生2》(The Sims 2)时的海量资产清单,他指出相比初代《模拟人生》,续作的内容量宛如一座庞大的高山。纵观他以往项目的开发团队规模演进,美术与内容制作团队的扩张速度远远超过了整体团队规模的线性增长,呈现出非线性的暴增趋势。 这种内容生产模式在经济学上是难以持续的:制作两倍的内容往往需要耗费两倍的开发精力,但带给玩家的效用(Utility)却远非翻倍。这就像经典的边际效用递减法则:当一个人流落荒岛濒临饿死时,得到第一个苹果意义非凡;但当他已经坐拥 100 个苹果时,第 101 个苹果就变得微不足道。 与之形成鲜明对比的是,玩家对亲手创作内容展现出了极其惊人的热情: - 在《模拟人生2》官方网站上,玩家自发上传了约 17,000 个角色和 10,000 座房屋。 - 大型玩家同好社区甚至刚刚举行了庆祝活动,纪念站点上第 100,000 个玩家自制物品的诞生。 - 哪怕这些 UGC(用户生成内容)中只有 5% 到 10% 称得上优秀,其总体规模也足以碾压开发团队人工制作的总和,更何况玩家作品的优质率还在持续走高。 此外,开发团队从玩家行为数据指标(Metrics)中发现,玩家极度反感游戏中的人造边界与空气墙,越来越渴望高自由度的开放世界。为了扩大游戏的“可能性空间”(Possibility Space),团队往往需要进行巨量的迭代开发。传统上,业内往往将玩家获得的价值简单视为“内容数量”与“内容质量”构成的二维面积——例如《皮克敏》(Pikmin)内容体量较小但美术品质极高,而《块魂》(Katamari Damacy)虽然画质粗糙但内容极其丰富,两者在二维价值梯度上提供了相近的游玩价值。 然而,Will Wright 尖锐地指出,这种认知忽略了一个极其关键的第三维度——“归属感”(Ownership)。一旦玩家拥有了对内容的创作与掌控权,哪怕内容数量不多、画质不精,玩家也会深度投入。为了说明这一点,他展示了自己在《侠盗猎车手》(Grand Theft Auto)里自创的角色“Mo”: - 这个角色不爱做主线任务,是个成天穿着可爱内裤、戴着芭比手表在街头练习自行车特技的无业游民。 - 每次进入严肃的主线过场动画时,Mo 的滑稽扮相都与严肃的黑帮剧情形成荒诞的对比。 - 游玩之后,Will Wright 会兴奋地跑去跟妻子和女儿讲述 Mo 今天的离奇遭遇,却从不会向家人复述游戏中预设好的剧情过场。 在总结归属感的重要性时,他坚定地断言:“我坚信玩家自身的故事永远比我们试图强加给他们的脚本故事更具感染力。”(“and so i contend that player stories will always be more powerful than scripted stories that we try to tell the players um”) ### [06:19](https://www.youtube.com/watch?v=ofA6YWVTURU&t=379s) 从苹果II到CD-ROM:算法的消亡与数据的统治 ```mermaid flowchart TD accTitle: 游戏内容存储与生产范式演变 accDescr: 展示从Apple II代码主导到CD-ROM数据主导的技术演变路径 subgraph AppleEra["Apple II 时代 (受限存储)"] direction TB A1["极其有限的软盘 (140KB)"] A2["内存代码与数据比例 1:1"] A3["贴近硬件的机器码编程"] A1 --- A2 --- A3 end subgraph Transition["存储媒介剧变"] CD["CD-ROM 引入 (超级油轮级的容量)"] end subgraph DataEra["数据至上时代 (预制内容膨胀)"] direction TB D1["代码与数据比例暴跌至 1:100"] D2["算法逻辑让位于海量预制资源"] D3["典型案例:《神秘岛》(HyperCard幻灯片堆叠)"] D1 --- D2 --- D3 end AppleEra --> Transition Transition -->|"撞击消灭算法优势"| DataEra ``` 结合内容成本激增、玩家生成内容的高价值以及开放游戏的大趋势,Will Wright 将演讲主题进一步具象为:《我从〈模拟人生〉中学到的内容,以及它为何驱使我走向过程化生成方法》(what i learned about content from the sims and why it's driven me to procedural methods)。 他回顾了自己在 Apple II 电脑上开始写代码的早期岁月。那台老机器在当年的广告中号称“比电传打字机快 6 倍”,开发者必须紧贴底层硬件编写机器码,人手一份详细的内存映射表,精确知晓每一块内存的具体用途。当时内存空间极其匮乏,内存中代码与数据的比例大致保持在均衡的 1:1,容不得挥霍哪怕一点点数据。而到了如今的次时代游戏开发中,这种代码与数据的比例已经悬殊地倾斜至 1:100。 打破这种技术平衡的关键转折点是 CD-ROM 的普及。相比以往容量只有 140KB 的小软盘,CD-ROM 就像是一艘装载量不可思议的“超级油轮”,其海量存储空间在当时完全超出了开发者的想象,引发了全行业对“到哪里找那么多数据塞满光盘”的困惑。这场媒介变革宛如一颗撞击地球的陨石,“在某种意义上它敲响了算法的丧钟”(“this basically was kind of like the meteor that came in and sounded the death knell for the algorithm”),自此开启了“数据至上、数据称王”的制作模式。 Will Wright 以当年风靡一时的冒险神作《神秘岛》(Myst)为例: - 该作由 Rand Miller 与 Robyn Miller 兄弟打造,本质上是基于 HyperCard 构建的一套极其精致复杂的“幻灯片系统”。 - 尽管技术实现形式只是预渲染幻灯片的切换,但它成功构筑出了极具沉浸感的世界,堪称划时代的杰作。 - 但在沉浸于游戏的同时,Will Wright 心中始终浮现出一个念头:创造这个世界的过程本身乐趣无穷,开发者在搭建这套世界时想必玩得无比开心,而这种亲手创造世界的极致快感,理应交到玩家手中。 ### [08:30](https://www.youtube.com/watch?v=ofA6YWVTURU&t=510s) 崇拜算法的北欧黑魔法:Demo Scene 的启示 当整个游戏工业都在走向以数据驱动为核心的道路——即试图用海量预制内容填满一张张 CD-ROM 时,世界的另一端却存在着一个恪守“古老法则”的群体。Will Wright 将他们形容为一个主要聚集在北欧、以赫尔辛基为中心的神秘部落,这群人被称作“Demo Scene”(演示圈)。与工业界大肆铺陈美术资产的做法相反,他们信奉算法的极致威力,在现代技术的掩盖下操练着数字时代的“黑魔法”。 这群狂热的极客能够仅仅利用 64 KB 极限大小的程序,完全通过纯粹的算法和过程化生成手段,凭空构建出极其宏大复杂的动画场景、视觉特效以及完整的音乐和音效。Will Wright 特别强调了 64 KB 的极限程度:“如果我们将这 64K 平方一次、再平方一次、再平方一次、再一次又一次,那才差不多达到一张 CD-ROM 的容量。”这个群体的技术演化源于早期的游戏破译社区(Cracking Community):破译者为了争夺荣誉,会在破解游戏的前端加入一段署名的小演示画面(Cracktro);随着技术较量日益激烈,这些开头动画变得越来越繁复华丽,最终演化成了一种定期聚会比拼技术与创意的黑客仪式。Will Wright 感慨这种将算法压榨到极限的探索在过去几年中给予了他无尽的灵感。 - Demo Scene 的发源地以北欧斯堪的纳维亚半岛为中心,核心腹地位于芬兰赫尔辛基。 - 极客团队使用 64 KB 极小体积的独立可执行文件,即可实时演算出媲美大型视听作品的声画效果。 - 这种创作文化从早期的软件破解署名动画演化为纯粹依靠数学与代码表现视听艺术的数字亚文化。 ```mermaid flowchart TD accTitle: Demo Scene Evolution accDescr: 展示从早期破解组竞争到算法极简演示的演变路径 subgraph Origin["起源阶段"] A["游戏破译组竞逐"] --> B["破解署名画面 (Cracktro)"] end subgraph Transformation["技术蜕变"] B --> C["追求极限视觉与音乐表现"] C --> D["摆脱预制数据,信奉纯算法与过程化"] end subgraph Culture["成熟文化"] D --> E["Demo Scene 算法极客群体 (64 KB 演示)"] E --> F["对 Will Wright 的过程化设计产生深远启发"] end ``` --- ### [10:18](https://www.youtube.com/watch?v=ofA6YWVTURU&t=618s) 显微镜的顿悟与组建“远古法则”特种部队 Will Wright 指出,本次演讲更准确的主题其实是:“我从《模拟人生》(The Sims)中学到了关于内容的什么经验,为什么它将我推向了过程化方法,以及我现在打算用这些方法来做些什么。”为了实现这种构想,他的第一步便是在业内招揽了一支精通这种“远古算法秘术”的精英开发团队。 在谈及开发背景前,Will Wright 聊起了一段看似离题却极为关键的生活插曲。多年前他的女儿五岁时,他打算同时买一台优质的天文望远镜和一台优质的显微镜。望远镜在各大科学商店或自然探索用品店随处可得,但寻遍大街小巷却买不到一台出色的显微镜,这让他深感惊讶。他极力建议同行在 eBay 上购置这类工具:“如果必须二选一,我一定会选显微镜,尤其是跟孩子一起玩的时候,这东西太酷了。”他特别补充了选购诀窍:必须挑选带有双目立体视觉(Stereo)和顶部照明光源的机型,而且反直觉的是,一定要买低倍率(Low Power)而非高倍率版本。只要把自己的手掌平放在镜头下,观察者就会沉迷其中长达一个小时,那种被微观细节震撼的体验宛如一次“宗教般的体验”。 - 天文望远镜代表朝向宏观尺度的远望,而低倍立体显微镜则带来了观察微观质感与细节结构的独特视角。 - Will Wright 随后将招募到的顶尖工程师们安置在一个处于隐蔽状态的现代化研发中心,秘密开启了新项目的攻坚。 --- ### [11:39](https://www.youtube.com/watch?v=ofA6YWVTURU&t=699s) 早期雏形公开:告别公关噱头的《孢子》(Spore)首秀 回到正题,Will Wright 宣布将当场演示团队的最新成果。他坦言自己平日里并不热衷于做实机 Demo,但这次之所以打破惯例,是因为相较于 E3 那种面向大众与营销导向的商业展会,他更希望首先把这种实验性的技术成果展示给自己深具归属感的开发者同行圈子。 展示在大屏幕上的成果在当时被称为《Spore》(孢子)。Will Wright 诚恳地提醒全场,当前展示的代码版本非常早期,“基本上是用捆包铁丝和胶带勉强粘在一起的”(held together with bailing wire and duct tape),距离最终形态还有极漫长的路要走。然而,这套粗糙的演示已足以清晰揭示出过程化技术之所以令人兴奋的本质原因。 > “在展示之前我想说,这还非常不完善,基本上是用铁丝和胶带捆绑在一起的,还有很长的路要走,但我认为我可以向你们展示足够的内容,让你们明白我为什么对过程化方法如此兴奋。” —— Will Wright --- ### [12:38](https://www.youtube.com/watch?v=ofA6YWVTURU&t=758s) 第一阶段实机:微观细胞世界的“流体吃豆人” 演示画面切入了一个处于“极微小尺度”(land of the very small)的世界。玩家操控着一个生活在细胞尺度环境中的单细胞微型生物,整个空间充满着实时的微观流体动力学演算模拟。Will Wright 形象地将这一阶段的游戏机制比作“一款极具模拟感、充满流体特性的吃豆人(Pac-Man)”。 小生物在流体中游弋并吞噬微小的食物浮游碎屑,同时还要躲避比自身更庞大凶猛的危险浮游生物。游戏的世界环境与生命体态几乎完全由过程化算法实时演算。当这只生物吞噬了足够数量的食物碎屑后,它便会排下一枚小小的黄色卵。点击这枚卵,游戏即刻切入世代进化的编辑器界面。 在初级编辑器中,玩家可以自由配置最基础的生物组件: - **口器构件**:影响进食方式,玩家既可以选择过滤浮游碎片的滤食口,也可以选择能够直接刺入猎物身体的口针(Proboscis)。 - **运动构件**:包含鞭毛(Flagella)、纤毛(Cilia)以及微型喷射器官(Jet)等不同推进机制。 - **自卫与捕食武器**:如刺针(Spikes)与刺细胞(Nematocysts)等。 Will Wright 在这只小生物的前端加装了一枚尖刺并返回模拟世界。原本能轻易猎杀它的褐色浮游生物,现在反而成为了它的猎物,进而打开了全新的食物链来源。随着生命的繁衍与身体结构的复杂化,虚拟摄像机镜头逐渐向上拉远,背景景深随之虚化,单细胞在数个世代的演进中不断成长扩张。 --- ### [14:26](https://www.youtube.com/watch?v=ofA6YWVTURU&t=866s) 3D 海洋与跨玩家的异步生态系统分发 经历多次世代迭代后,游戏环境从原本的二维微观水滴无缝跨越到了三维的海洋深处。此时的整个海底世界不仅是纯粹依靠算法实时生成的,甚至连玩家所操控的生命体,其骨骼绑定、动画位移模式、皮肤质感和反应行为也全由过程化系统自动推算赋予。 更为颠覆性的机制在于生态系统的构成来源。Will Wright 指出,深海中游弋并与玩家角逐生存空间的庞杂生物,并非来自光盘预装的数据,更非开发团队预先摆放的 NPC: > “当我说这些生物来自其他玩家时,我的意思是它们正在被异步发送到你的电脑上以充实你的世界;我们实际上是在根据你所在世界的需求,构建一个平衡的生态系统。” —— Will Wright 这套机制利用网络在后台将全世界玩家创建的生物数据进行异步下载分发,依据单机当前的食物链层级与生态缺失,动态填充出丰富而真实的本土生物圈。 ```mermaid flowchart LR accTitle: Asynchronous Ecosystem Distribution accDescr: 展示玩家创造的内容如何通过云端异步分发并动态维持本地生态平衡 subgraph UserA["玩家 A 客户端"] A1["在编辑器创造新物种"] --> A2["上传物种轻量数据"] end subgraph Cloud["后台生态平衡服务"] A2 --> B["分析物种属性与生态位"] B --> C["匹配全球玩家的世界需求"] end subgraph UserB["玩家 B 本地沙盒"] C --> D["异步下载物种数据"] D --> E["过程化解析骨骼、动作与材质"] E --> F["填入本地生态链缺失环节"] end ``` --- ### [15:52](https://www.youtube.com/watch?v=ofA6YWVTURU&t=952s) 自由造物与三足物种的陆地拓荒 深海生物完成繁殖后,再度产下受精卵。点击卵后,演示展现了高阶生物编辑器的核心技术。Will Wright 解释,这个编辑器的操作哲学融合了“土豆头先生”(Mr. Potato Head,自由插拔五官构件)、“组合积木”(Erector Set,机械式结构组合)与“雕塑粘土”(Clay,自由塑形表面)的综合特征: 1. **骨骼与体态塑形**:生物具备一条可视化的基础脊柱,拉伸或弯曲脊柱不仅会直接改变骨骼骨架,利用鼠标滚轮还能像捏软泥一样自由缩放局部躯干的粗细,从而光滑地重塑外形。 2. **功能性组件装配**:组件依照功能系统分类,涵盖抓取器官(Graspers)、攻击武器、推进足部等。许多配件兼具复合属性,例如某些抓取爪本身就带有轻微的攻击伤害力。 3. **肢体自由装配与尺寸放缩**:玩家可以向身体任意部位粘接不同关节数量的肢体。Will Wright 摘掉了原生物用于游泳的鳍,换上了一个三关节腿与脚掌,并额外加装了一条腿,将其改造成一只独特的“三足生物”(Tripod)。 当这只丧失了鳍片、长出三条腿的怪异生物重返游戏时,它在水中的游动能力大打折扣。Will Wright 索性操控它向海滩移动,完成了生命登上海陆交界处的历史性瞬间。 最令全场惊叹的技术细节在于运动步态的自适应合成:玩家在编辑器内无论拼凑出多么畸形或者反常的身体形态,游戏底层算法都会基于对其新脊椎和骨骼构架的受力与拓扑分析,完全通过过程化实时演算出这只三足怪物的爬行步态。游戏内的每一个动词交互(Verbs)均无需预制动画资产,而是全部依托算法赋予这些生命鲜活的动态反应。这只刚刚走出海洋的三足生命,即将面对由其他玩家创造的全新陆地物种所交织出的辽阔世界。 ### [18:22](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1102s) 生物生存、程序化动词与危险的生态系统 Will Wright 控制着刚刚创造好的生物在三维世界中活动。由于这只生物被设定为食肉动物(carnivore),它必须通过捕食其他生物来维生。Wright 指挥它靠近并攻击场景中的另一只小生物。生物在战斗中的动作直接取决于玩家把武器安置在身体的哪个部位——因为这只生物的攻击武器被安装在尾巴上,必须调整好距离才能打中;如果距离不够远,攻击就会落空。所幸 Wright 的生物速度比猎物更快,在一段追逐后顺利击杀了目标,甚至逗趣地喊着让对方从自己头上滚开。击杀之后,生物便开始啃食尸体。 在此处,Wright 着重展示了“程序化动词”(procedural verbs)的强大之处:它们可以极其自然地进行混合。例如,玩家可以对生物发出指令让它咬住猎物,同时又下达移动行走的指令,系统便会自动将这两种基础动作融合成一个全新的衍生动词——“拖拽”(drag)。Wright 强调,这种基于玩家构建的生成机制带来了本质上的行为差异,正如他所评价的:“无论我在这个游戏里创造了什么,它的行为方式都会截然不同;我的意思是,它在本质上就会和你在这款游戏里可能创造的任何东西有着全然不同的行为,这是一个巨大的胜利。”(“one of the real wins though is that whatever i've created in this game is going to behave differently i mean you know it's just going to be fundamentally different behavior than anything that you might create in the game uh which is a huge win”) 就在 Wright 示范拖拽时,视野中出现了一只体型庞大且危险的敌对生物。Wright 立刻高呼“三十六计走为上策”(run away)。游戏内的算法会动态构建一个平衡的生态系统,玩家永远不会自然而然地稳坐食物链顶端。面对体型巨大、移动迅速的捕食者,Wright 灵巧地改变走位,引诱怪物去追赶那些蹦蹦跳跳的弱小生物,成功化险为夷。 ### [20:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1221s) 求偶、繁衍与生物形态的多样性展示 在野外成功生存并度过一段生命周期后,生物需要通过繁殖进入下一阶段。在当前游戏阶段,求偶交配不仅能为玩家赚取货币,也是解锁生物编辑器的唯一途径。Wright 按下按键让生物发出求偶鸣叫(mating call),在原野中搜寻配偶。他小心翼翼地靠近回应呼唤的同类,双方随即触发了一段诙谐的程序化求偶动画,引得全场哄堂大笑并爆发掌声。完成交配后,生物产下一枚蛋,系统随即进入生物编辑器(Editor),让玩家为下一代生物定制进化路线。 为了向台下开发者直观展示编辑器的自由度与开放性,Wright 依次加载了开发团队在编辑器中预先制作的几个典型生物: - **Buttface(屁股脸)**:长着一副极不寻常的腿部骨骼拓扑结构,但运动系统依然能驱动它平滑、逼真、栩栩如生地奔跑。Wright 幽默地解释了一个心理学把戏——正因为人们现实中从未见过“屁股脸”,所以大脑根本不知道它“应该”怎么走,反而不会像看人类行走那样挑出破绽。 - **Tweety Bird(崔弟鸟)**:顶着一颗比例夸张的巨大脑袋。Wright 戏称它的力学结构就像一台高重心的 SUV 或福特探险者(Ford Explorer),跑动时前倾后仰随时要翻车,并调侃“我怀疑它在现实世界中的适应度会非常低”(“i suspect he would have low fitness in the world”)。 - **Hang Dog(垂头狗)**:外观与步伐都极其冷峻帅气。 - **Dilemma(进退两难)**:步态频率极快。它的行动表现既取决于所选身体部件,更取决于玩家搭建的拓扑骨骼结构,程序化动画与骨架相互作用从而涌现出了超凡的奔跑极速。 - **Care Bear(爱心熊)**:这是 Wright 强烈要求团队做进系统的角色。他打趣道,自己偏要让这只长相可爱的爱心熊变成极其残暴的肉食怪兽。 Wright 强调,所有这些外形千奇百怪、走路姿态截然不同的生物,底层其实跑的都是同一套分析骨骼并赋予生命的程序化驱动脚本;未来游戏还将融入完整的程序化纹理着色流水线。 ### [24:01](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1441s) 游戏内嵌创作、心智归属感与“爱心熊宇宙” 回顾《模拟人生》(The Sims)的创作生态,Wright 指出过去最大的痛点在于玩家必须跳出游戏,单独下载外部工具制作内容,然后再上传到专门的网站,流程极度繁琐断裂。而在《孢子》中,核心设计目标是将编辑器做成内嵌在核心玩法中的“玩具”(toy),让内容在玩家之间的流转变得完全无缝透明——玩家无需四处寻找并下载外星球生物,系统会分析玩家的游玩方式,全自动将其他玩家创造的匹配生物拉取到其星系中。 针对以往进化类游戏中玩家缺乏“归属感”的通病,Wright 批判那些作品往往只提供“头部、躯干、尾巴”三段式插槽供人机械挑选。他回忆起自己童年对恐龙的痴迷:“在学习不同恐龙类型和名字的过程中,我很得意自己知道什么是甲龙(Ankylosaurus),而我妈妈却不知道,这让我对它们产生了一种独特的拥有感。”无论是过去的恐龙,还是当代的宝可梦(Pokémon)或尼奥宠物(Neopets),孩子们一旦掌握了其背后繁复的规则与属性体系,就会获得强烈的掌控欲和主人翁意识;而如果这种生物是由他们亲手从零创造的,那种心智连接将无与伦比。 紧接着,Wright 借着爱心熊(Care Bears)展开了一段令人捧腹的叙事发散: - 爱心熊最初不过是美格立(American Greetings)用于贺卡的角色,用来对应不同情绪; - Wright 坦白自己曾在网络心理测试中测出自己对应的角色是“日光熊”(Sunshine Bear); - 后来官方又发明了不是熊的“爱心熊表亲”(Care Bear Cousins),它们生活在“感情之林”(Forest of Feelings),而爱心熊生活在云端的“爱心之乡”(Care-a-Lot); - 当小熊从云端掉落时会掉进感情之林,而地球还在那更下方——这意味着小熊下凡执行任务时,其居住地大概位于“近地轨道”(low earth orbit); - 官方将这片领地统称为“关爱王国”(Kingdom of Caring),而 Wright 戏谑道:“他们虽然从没挑明,但话里话外都在暗示,地球其实是‘谁在乎王国’,因为那里是他们不得不跑去出任务的鬼地方。”(“and they never specifically say it but they kind of imply that earth is the kingdom of who gives a because that's that's where they always have to go on their missions”) ```mermaid flowchart TD accTitle: Care Bears Lore Cosmology accDescr: Diagram depicting Will Wright's satirical breakdown of the Care Bears universe hierarchy down to Earth. subgraph Heaven["关爱王国 (Kingdom of Caring)"] CL["爱心之乡 (Care-a-Lot)<br/>爱心熊居所 / 近地轨道"] FOF["感情之林 (Forest of Feelings)<br/>表亲居所 (落云缓冲层)"] CL -->|"自云端跌落"| FOF end subgraph Mundane["凡间执行区"] Earth["地球 (Earth)<br/>被调侃为'谁在乎王国' (Kingdom of Who-Gives-A-*)"] FOF -->|"下凡执行任务"| Earth end ``` ### [26:35](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1595s) “盒中艺术家”与 5000:1 的创意放大器 在谈及开发理念时,Wright 引用了首席美术师 Ocean 提出的概念——“盒中艺术家”(artist in a box)。团队系统梳理了传统游戏美术师产出资源时的所有繁复步骤,思考如何将其全盘自动化,从而只给玩家留下一套极度简化的控制手柄,让普通玩家感受到巨大的“创意杠杆”(creative leverage)。在视觉美学风格的光谱上,游戏旨在横跨皮克斯(Pixar)的圆润童趣与 H.R. Giger(异形之父)的暗黑机械生物风,既能孕育苏斯博士(Dr. Seuss)式的奇幻世界,也能诞生硬核逼真的恐怖异形。 在底层系统流水线中,玩家仅仅是在操作一套极其轻量的结构: ```mermaid flowchart LR accTitle: Creative Amplification Pipeline accDescr: Pipeline showing how simple player actions are expanded procedurally into rich assets. subgraph Input["玩家微型输入 (~1 KB)"] P["玩家直观编辑<br/>简易骨架与拓扑"] end subgraph Engine["盒中艺术家 (程序化放大器)"] Mesh["网格拓扑解析"] Tex["程序化纹理脚本"] Anim["程序化骨骼行走动画"] AI["行为系统 (自驾 / NPC 驱动)"] end subgraph Output["最终游戏呈现 (~5 MB)"] World["逼真鲜活的世界实体"] end P --> Mesh Mesh --> Tex Mesh --> Anim Anim --> AI Tex --> World AI --> World ``` 计算机在这一过程中充当了关键的“创意放大器”(creative amplifier)。数据压缩与放大层面的表现尤为惊人: - **原始数据规模**:玩家在编辑器内所做的全部结构与参数定义,经过序列化后仅占用约 **1 KB** 的内存空间。 - **运行时展开规模**:当引擎动态生成出完整的网格(mesh)、程序化材质纹理(procedural textures)、骨架动画序列以及行为驱动脚本后,其总体数据量约为 **5 MB**。 - **放大倍率**:这达成了惊人的 **5000:1** 膨胀/压缩比。 这种极端的数据轻量化带来了颠覆性的工程收益:玩家获得了撬动复杂动画与形体的超级杠杆;海量异星生物可以以微乎其微的体积储存在轻型数据库中,即使在玩家断网离线的情况下,游戏也能轻松塞进一个生态丰富、物种浩瀚的生动宇宙。 ### [28:39](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1719s) 数据压缩与授粉系统:程序化生成的基因型与表型 在离线状态下,游戏光盘自带的数据库就已预装了成千上万种生物;而一旦连接网络,极小的数据体积让这些内容在玩家之间(无论是点对点 P2P 还是客户端-服务器架构)分发的摩擦力降到了极低。威尔·赖特(Will Wright)指出,赋予玩家强大创造力的底层杠杆,与构建程序化生成系统所需要的机制完全相通——计算机完全可以像玩家拨动滑块一样去操纵这些参数。这正如同在《模拟人生》(The Sims)中捏脸或定制角色,玩家调整滑块,系统通过内部流程生成最终资产。 这种机制在本质上完美映射了生物学过程:数据本身如同“基因型”(Genotype),经由类似“胚胎发育”(Embryology)的算法解压过程,最终转化为游戏世界中可见的“表现型”(Phenotype)。威尔·赖特在此强调:“and this is kind of basically mapping to the idea of a genotype going through embryology to an initial phenotype and that compression is vital for this whole thing.”(这基本上映射了基因型通过胚胎发育转化为初始表现型的理念,而这种压缩对整套体系来说至关重要。)依托这种极度精炼的数据结构,游戏实现了名为“授粉”(Pollination)的机制:玩家创造的生物上传至中央服务器,服务器根据其生命周期、行为习性、体型大小、食性等参数进行分类与优选匹配;当玩家在单机世界中冒险并需要竞争物种来充实生态位时,系统便会自动抓取其他玩家创造的生物填充进来,在游玩过程中源源不断地动态构建生态圈。 ```mermaid flowchart TD accTitle: 孢子授粉与内容生成流程 accDescr: 展示生物数据从玩家创造到服务器分类,再到下载至其他玩家生态系统的循环过程。 subgraph ClientA ["创造端(玩家 A)"] A1["生物编辑器参数<br>(基因型 Genotype)"] --> A2["算法解压/程序化生成<br>(胚胎发育 Embryology)"] A2 --> A3["动态生物实体<br>(表现型 Phenotype)"] end subgraph Cloud ["中央分发体系(授粉 Pollination)"] B1["轻量化数据上传<br>(P2P 或中央服务器)"] B2["参数化标记与分类<br>(体型/食性/行为/生命周期)"] B1 --> B2 end subgraph ClientB ["接收端(玩家 B)"] C1["本地生态系统发出填充需求"] C2["匹配下载匹配物种"] C3["实时动态构建竞争生态"] C1 --> C2 --> C3 end A1 -.->|"极小数据体积分发"| B1 B2 -.->|"参数映射拉取"| C2 ``` - **数据轻量化优势**:仅需传输参数指令而非庞大网格模型,光盘原生容纳数万生物,网络传输近乎零延迟阻力。 - **参数化生态匹配**:服务器不仅按品质筛选热门模型,还深度分析生物的行为逻辑、食物链层级与生态功能,确保动态拉取的物种能够合理维系生态平衡。 --- ### [30:03](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1803s) 脑力升级与部落阶段:从单体进食到“拿枪的猴子” 在生物编辑器最下方设有一个特殊的“脑力面板”(Brain Palette)。在单体生存阶段,玩家往往习惯于在受到袭击时把点数用于购买更好的防御武器、伪装或奔跑速度,但真正明智的长期策略是像打理“大学学费储蓄账户”(College Savings Account)一样,逐步将资金沉淀并投资于大脑容量。随着大脑梯度的逐级提升,物种最终突破临界点觉醒智慧,游戏机制也在此刻发生基因突变——从原先第一人称视角的“大鱼吃小鱼”吞食模拟,无缝蜕变成为即时战略(RTS)模式,玩家开始统领一整群生物。 威尔·赖特谈到他在游戏设计中念念不忘的标志性意象(Landmark):“one of the landmarks is I've always been fascinated with the idea of monkeys with guns.”(我一直着迷于‘给猴子一把枪’这个概念。)在这一阶段,生物的肉体进化已经锁定,核心玩法转变为社会化与工具研发。赖特在现场演示了为生物族群购买工具:建造“矛架”(Spear Rack)后,族群表现出高度动态的程序化动作反应与互动欲望;为了防止它们惹是生非,赖特又购买了一座“营火”(Campfire),以此调控氏族的社交动态与情绪机制,使族群根据玩家的引导在“高度情绪化”与“高度理性化”之间演化。随后他为部落放置了一面鼓,生物们不仅主动敲击乐器,更围绕营火跳起了基于其独特身体结构的“程序化舞蹈”(Procedural Dance),这一极具生命力的表现引来全场掌声。 - **玩法流派演变**:游戏体验在此处向牛蛙经典神作《上帝也疯狂》(Populous)靠拢,玩家通过工具投放引导群体行为,指派狩猎队外出打猎并与其他部落竞争。 - **部落发展核心指标**:升级族群的“中心草屋”(Hut)是部落阶段的核心晋升阶梯;草屋等级不仅决定了高阶工具的解锁,还直接制约着部落人口上限与食物资源维系能力。 --- ### [33:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=1984s) 城市阶段与建筑编辑器:乐高式拼装与跨玩家审美共享 随着部落草屋升至顶级,游戏平滑迈入“城市阶段”(City Game)。该模式在机制上相当于一套精巧极简版的《模拟城市》(SimCity):玩家规划道路、建造房屋、解决市民的各项需求,维系市民的幸福感。在该阶段,玩家对视野中的每一栋建筑拥有百分之百的创意创作权。城市商店目录中陈列的建筑同样来源于网络上其他玩家的创作;算法甚至会智能学习玩家当前的建造偏好与审美流派,定向推送风格契合的作品供玩家一键购买。 如果玩家选择从零打造或修改建筑,即可进入与生物编辑器逻辑完全相通的“建筑编辑器”。相较于生物的骨骼与肌肉形变,建筑编辑器在体验上更接近乐高积木(Legos):系统提供立方体、球体等极简的正多面体或柏拉图立体(Platonic Solids),每个图元都带有操纵手柄,支持自由形变、相交穿插与垂直堆叠。赖特回忆道,早在《模拟人生》的建造模式中他们就发现,尽管建造工具存在轻微的学习曲线,但玩家一旦上手就能获得极强的即时正反馈与创造性成就感,因此团队致力于将这种纯粹的创造乐趣更深地缝合进核心流程中。 在展示中,赖特为市政厅添加了一个蘑菇状的奇异结构,返回游戏后市民们立刻展现出欢欣雀跃的正面情绪。随后他将镜头平移横跨地图,展示了一座由算法抓取而来的邻近电脑城市:两座城市均出自同一套编辑器,但赖特自己的城市呈现出异想天开的“苏斯博士风格”(Dr. Seuss Style),而邻邦则是通体硬朗冷酷、带有重度机械质感的“星战科技工业风”(Techno-industrial Star Wars Style)。这种跨度鲜明印证了系统能兼容从“皮克斯”(Pixar)式童趣到“吉格尔”(H.R. Giger)式异形工业美学的超广谱创意区间。 --- ### [36:20](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2180s) 载具建造与文明演进:三类同盟手段与全球视野 当城市人口与基建升级到特定规模后,系统将解锁“载具编辑器”(Vehicle Editor)。在部落阶段,势力间的交互依托于派出个体进行突袭掠夺或建立贸易;而在城市阶段,城邦之间的征服与结盟分化为三条截然不同的路径:军事手段(Military)、经济渗透(Economic)或文化同化(Cultural),具体取向完全由玩家此前培育的族群文化特征决定。 载具编辑器复用了建筑编辑器的操作逻辑,但其组件开始具备严谨的功能学属性:螺旋桨赋予飞行能力,特定底盘与构件决定水面航行或陆地机动能力。演示中,面对邻邦极具攻击性且正向己方推进的工业化军队,赖特临时选购了一架外形滑稽、带有纳斯卡(Nazca)地画风格的自制奇特载具迎战。然而实机操作中,这支自制战车却呈现出令人捧腹的混乱 AI 反应——不仅完全不听指令转向逃窜,还被赖特当场调侃为“爱好和平的胆小鬼”。 随着地缘冲突从双城对峙进一步扩大,游戏正式过渡至宏大的“文明阶段”(Civilization Game),机制演变为更宏观维度的简化版《文明》(Civ)或桌面战棋《战国风云》(Risk)。当摄像机视口在此刻首次大幅度向外拉升,观众席得以窥见一个由六边形地块构成的完整三维行星球体,数座各具文明特色、拥有不同意识形态的城邦在星球表面星罗棋布。 - **城邦交互的三大支柱**:军事征服(武力推平)、经济收购(商贸垄断)与文化传导(意识形态归化),对应不同的载具配置与战术重心。 - **组件功能驱动**:载具编辑器的部件并非纯外观皮肤,推进器、浮力装置、轮轨等不同模块直接决定了单位的物理通行域(陆海空全域机动)。 ### [38:10](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2290s) 城市尺度的无缝缩放与内容授粉机制 Will Wright 在现场演示中展示了视角拉远时的视觉变化:当镜头远离城市时,城市建筑的外观会被刻意夸张化。这种设计兼顾了美学表现与功能需求。当镜头推近时,城市会切换至逼真的真实尺度;而一旦镜头拉出,整座城市就会缩略成一个图标化的交互单元,直接指示出当前微观管理的层级。“在这个层级上,城市就是一个大按钮,拥有一个非常简单的菜单;如果我想进去继续改造城市,比如在敌人进攻的同时继续建造‘巴黎’,我依然可以放大视角,点击单个建筑物并把它们调入编辑器等等。”在《文明》(Civilization)式的阶段中,玩家一步步攻城略地、占领大陆,直到彻底掌控并建设整颗星球。 ```mermaid flowchart TD accTitle: 玩家生成内容与中央数据库的授粉与排序机制 accDescr: 展示其他玩家创造的资产如何推送到购买模式,以及玩家的购买行为如何反哺优化主数据库的资产排序。 subgraph ClientSide ["客户端体验"] A["玩家当前游戏风格"] --> B["购买模式 (Buy Mode)"] B --> C{"选择资产来源"} C -->|"自主设计"| D["调用内置编辑器创作"] C -->|"直接采购"| E["购买其他玩家的内容"] end subgraph ServerSide ["中央数据库 (Master Database)"] F["全网玩家内容池"] -->|"参数化匹配"| B E -->|"正向反馈信号"| G["统计全网使用次数"] G -->|"持续重排序 (Resorting)"| F end ``` 随后,Will Wright 深入拆解了该阶段背后的“授粉”(Pollination)机制。在游戏的许多关卡中都设有“购买模式”(Buy Mode),系统会将其他玩家创造的内容通过参数化匹配的方式,推送给符合其游玩风格的玩家。玩家既可以亲自设计,也可以直接购买他人的作品。每一次购买行为都会作为一次正向反馈回传至中央数据库;系统根据全网实际使用该资产的玩家人数,对海量内容进行持续的重排序与筛选,从而不断挖掘出最优质的内容资产。 --- ### [39:48](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2388s) 跨类型致敬与轻量化核心玩法设计 谈及整体设计理念,Will Wright 指出《孢子》在某种程度上与任天堂的《瓦里奥制造》(WarioWare, Inc.)有着异曲同工之妙。他坦言自己希望借这部作品向自己最喜爱的众多策略与经典游戏致敬,尽管游戏行业有一条人尽皆知的准则——“切忌混搭游戏类型”(Don't mix genres),但他直言自己一直极度渴望打破这项规则: - **潮池阶段(Tide Pool)**:体验类似于《吃豆人》(Pac-Man)。 - **生物进化阶段(Evolution)**:体验反直觉地神似《暗黑破坏神》(Diablo)——深入危机四伏的世界砍杀怪物,侥幸生还后返回安全营地换取更精良的护甲与武器,再外出挑战更强大的怪物。这一循环完美契合了“进编辑器改造—活过一代—再返回编辑器”的代际交替体验。 - **部落阶段(Tribal)**:致敬 Peter Molyneux 的经典神作《上帝也疯狂》(Populous)。 - **城市阶段(City)**:毫无疑问致敬了《模拟城市》(SimCity)。 - **文明阶段(Civilization)**:融入了席德·梅尔的《文明》精髓。 Will Wright 强调,他并不想把系统做成动辄耗费几十小时的重度大作,而是力求提炼核心:“我希望它只保留《模拟城市》最精华的20%,或者《文明》最精华的20%。”他解释道,作为硬核玩家,自己现在根本没有那么多时间沉浸在繁重的大作中,反而花大量时间玩掌机 Game Boy。他希望《孢子》能将核心要素高度提炼,预计约 40% 的体验属于纯粹的美学与创意搭建,让玩家自如打造世界,而底层系统则保持规则极简、易于上手却富有深度的轻量化策略。每个阶段均配备专属编辑器供玩家核心交互。更为关键的是,角色的定位发生了根本性转变:“我不想把玩家放在卢克·天行者或佛罗多·巴金斯的角色上;相反,我想让玩家扮演乔治·卢卡斯、J·R·R·托金或苏斯博士。” --- ### [41:49](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2509s) 叙事哲学:故事是体验的副产物而非前提 针对游戏剧情的设计,Will Wright 阐述了一种截然不同的叙事视角。他认为剧情不应作为游戏的前提条件,而是有趣体验所自然生成的“副产物”或固有属性。他以《侠盗猎车手》(Grand Theft Auto)为例说明:“我甚至不知道《侠盗猎车手》原本的故事是什么,但我自己的故事非常有趣;我玩得很开心,而且这对我来说意义非凡——我拥有了那个故事。”他承认这并非《星球大战》或《哈姆雷特》式的文学巨作,但却是玩家完全属于自己的独特记忆。 在 Will Wright 看来,故事扮演着经历过滤器的角色,其核心功能在于强化因果关系。电影导演可以精准把控镜头,向观众呈现某人为何发怒、为何踢飞物品进而折断脚趾;游戏开发者虽然无法在流程中拥有这种全知视角的绝对控制权,但玩家却始终身处体验的核心,直接承受事件的因果。此外,故事具备便于交流的“可移植性”,能够将一段时空的经历转述给他人;同时它也是一种“压缩与抽象化”机制,使人类能够提炼出更高阶的认知图式并应用于不同场景。他援引罗伯特·路易斯·史蒂文森创作《金银岛》的轶事指出:史蒂文森最初只是画了一张海岛地图,随着视线在地图上流淌、脑海中构想各种可能发生的情境,故事便水到渠成地自然涌现了出来。 --- ### [43:29](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2609s) UFO 工具与星际视角的宏大拉伸 视角随后切换回星球表面那个犹如精巧发条玩具的小世界。重返之前的城市后,地面上还残留着建筑被炸毁后留下的弹坑。在载具编辑器中,随着游戏进程逐步解锁地面、水面、空中及水下载具后,玩家最终解锁了终极移动载具——UFO。Will Wright 在现场直接购买了一架基础版 UFO,并指出这架飞碟不仅是一艘飞船,更是玩家在游戏后期集各种能力于一身的“瑞士军刀”,速度极快且机动性极高,片刻之间即可掠遍全球。 ```mermaid flowchart LR accTitle: 从地表微观到星系宏观的无缝视差变焦 accDescr: 展示 UFO 载具解锁后,镜头从城市地表持续拉远直至星系沙盒的宏大体验。 A["地表城市与生态环境"] -->|"UFO 绑架光束抓取物种"| B["行星宏观全貌"] B -->|"镜头持续向外拉远"| C["恒星系统恒星/彗星"] C -->|"跳跃至冰冷外缘行星"| D["外星荒原沙盒/改造改造"] ``` 谈到 UFO 的功能灵感,Will Wright 穿插了一段童年趣事。他小时候极度痴迷科普节目《奥马哈互助野生动物王国》(Mutual of Omaha's Wild Kingdom),主持人马龙·珀金斯(Marlon Perkins)等人每周都要深入野外拯救或研究濒危动物。令他印象最深的是,节目中几乎每一次解决生态问题的方式,“几乎总是包含一把高威力麻醉枪和一架直升机”。他每次守着电视就是为了看结尾直升机追逐兽群、探员挂在机外精准射击的刺激场面。因此,他将这种捕获机制移植到了 UFO 上——飞碟配备了“绑架光束”(Abduction Beam),既是一个探索收集工具,也是内容创造工具。在现场演示中,Will Wright 驾驶 UFO 飞向早期阶段曾欺负过自身物种的敌对生物,戏称“现在轮到我们复仇了”,随后精准开启光束将其中一只吸入飞舱。 紧接着,Will Wright 操纵 UFO 开始向太空加速爬升,镜头随之不断拉远、拉远、再拉远——伴随着全场观众雷鸣般的掌声与欢呼,整颗家园星球骤然缩小为星系中的一颗圆球,展现出完整的恒星系全貌。视界中呈现出彗星掠过的轨道、中央散发光芒的太阳、内侧的高温行星以及外缘冰冷死寂的星球。每一个天体都是一个供玩家施展创意的开放沙盒。Wright 随即选定了一颗地质风貌迥异于母星、远离恒星的严寒卫星,驾驶 UFO 俯冲降落至其冰冷荒芜的地表。他表示希望通过这套系统“引导生命诞生”(Bootstrap up to life),从零构建一个繁荣的全新文明。然而,当他在死寂的冰原上打开舱门释放刚才抓获的外星生物时,生物刚一触地便瞬间暴毙瘫倒在地。Will Wright 随即幽默地向台下观众发问:“有人能告诉我刚才为什么会发生这种事吗?”并示意现场调音台推高音频音量。 ### [47:43](https://www.youtube.com/watch?v=ofA6YWVTURU&t=2863s) 行星地形改造与殖民地建设 在抵达一颗荒凉的行星后,Will Wright 发现这里完全没有大气层,根本无法维持生命。为了让这颗星球变得宜居,他引入了行星地形改造(terraforming)系统。在设计这套系统时,他的灵感来源于经典的儿童绘图软件 Kid Pix。Wright 回忆起当年玩 Kid Pix 的乐趣,正是因为里面充满了稀奇古怪、极具创意的工具,而且玩家经常会失控搞砸,“你总会干出那种意料之外的‘哎呀,我不是故意的’这种事”,他希望地形改造也能带来这种不可预测的创造乐趣。 在演示中,Wright 激活了 UFO 的火山工具,直接在行星表面雕刻地形并向外喷发气体,逐渐为星球注入大气。但改变整颗星球的气候需要时间,在等待大气层形成的过程中,玩家已经可以降落建立殖民地。初期建立的殖民城市带有透明护罩(Bubble),这种穹顶防护罩技术同样适用于母星的海底城市建设。殖民城市能为玩家赚取资金,用于进一步升级 UFO 的各项机能。为了加快演示进度,Wright 使用了游戏后期才能购买的高级作弊装备——“创世纪装置”(Genesis Device)。装置启动后,地表质感迅速泛起绿意,暴风云散去,大气层逐步建立,生物圈得分(Biosphere score)随之提升。随后,玩家便可以把母星采集来的植物移植过来,让这颗星球真正繁衍起来。 - 地形改造工具通过 UFO 升级阶梯逐步解锁,使用火山等工具可以直接改变地表形态并释放大气。 - 在缺乏大气的星球或母星海洋深处,玩家均需建造带有防护穹顶(Bubble cities)的基地。 - 殖民地不仅扩大了文明疆域,还源源不断产生经济收益,构成了“建立殖民地赚钱 → 升级 UFO → 探索改造更多星球”的游玩循环。 - “创世纪装置”(Genesis Device)为高等级终极工具,可极速恢复大气并激活行星生态。 ```mermaid flowchart LR accTitle: 行星改造与殖民循环 accDescr: 展示了从荒凉行星到繁荣生物圈的改造与经济升级循环 subgraph Terraforming["行星宜居化"] A["荒凉无大气行星"] --> B["火山工具喷发气体"] B --> C["创世纪装置加速"] C --> D["建立生态圈与植被"] end subgraph Economy["经济与扩张"] D --> E["建设殖民城市"] E --> F["产生资金收益"] F --> G["升级 UFO 科技能力"] G -->|"开启星际探索"| A end ``` --- ### [51:00](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3060s) 星际跃迁与立体宇宙动物园 当玩家积累了足够的财富并不断升级 UFO 后,便能解锁极为昂贵的“星际跃迁引擎”(Interstellar Drive),从而跃出当前恒星系,进入真正的银河尺度。Wright 称这个辽阔的宇宙空间为“恒星动物园”(Stellar Zoo)。他尖锐地批评了传统天文学教育与星图的表现方式:教科书总是把宇宙绘制成平面的二维星图,让人完全无法体会宇宙真实的宏大与深度。“我们有来自哈勃望远镜的这些绝美照片,但我们从来不知道它们有多大、处于什么位置、在太空中出现的频率有多高”,Wright 表示,他极度渴望在这个阶段给玩家带来置身三维银河系中的真实震撼感。 在由程序生成的 3D 星系中,玩家可以探索各种真实的天文奇观:吞噬一切的黑洞、垂死恒星形成的行星状星云(Planetary Nebula,即超新星遗迹),以及孕育新生恒星的神秘星云。有些极其明亮的蓝巨星(Blue Giants)甚至即将发生超新星爆发,玩家在航行时必须谨慎远离,这让整个三维宇宙空间拥有了如同地形一般的危险起伏。此外,玩家还可以启用 SETI(搜寻地外文明)雷达工具,将鼠标悬停在不同恒星系的行星上监听无线电信号,搜寻地外智慧生命的踪迹。 - 星际引擎是 UFO 进阶路线中成本极高的关键升级,赋予玩家跨星系航行能力。 - 宇宙空间包含黑洞、行星状星云以及处于爆发边缘的危险蓝巨星,构建了立体的空间地貌。 - 玩家可利用 SETI 工具扫描并监听行星的电磁通讯,若截获外星语广播,即可确认该星系存在智慧生命。 --- ### [52:41](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3161s) 遭遇外星文明:跨玩家数据共享与星际冲突 当监听雷达截获外星广播后,Wright 驾驶飞船降落至目标恒星系的一颗靠近恒星、环境炎热干燥的行星上。这标志着玩家正式进入了由其他玩家创造的世界。Wright 郑重向全场解释了其背后的技术架构:“我想强调的是,这是异步的,它不是一个同步的多人游戏;我是从其他玩家那里下载了数据,然后电脑根据该玩家进化该物种时设定的性格滑块来接管控制。” ```mermaid flowchart TD accTitle: 异步多人生态交互架构 accDescr: 说明玩家创作的数据如何被云端抓取并在其他玩家单机中由 AI 驱动 subgraph Creator["玩家 A (创作者)"] A1["设计生物与载具"] --> A2["设定性格偏好滑块"] A2 --> A3["上传至中央数据库"] end subgraph Consumer["玩家 B (探索者)"] B1["SETI 监听定位星球"] --> B2["异步下载世界数据"] B3["本地 AI 接管外星文明"] B2 --> B3 B3 --> B4{"第一次接触机制"} B4 -->|"音调沟通"| B5["结盟或建交"] B4 -->|"谈判破裂"| B6["空对地武器交火"] end A3 -.->|"后台数据授粉"| B2 ``` 当飞船缓缓靠近地表一座繁华的外星都市时,游戏进入了“第一次接触”阶段。该环节借鉴了电影《第三类接触》的形式,双方需要通过一套音乐音调游戏来尝试建立共同语言。然而,Wright 在演示中沟通失败,外星防空火力瞬间全开向 UFO 倾泻子弹,场面立刻变成了《太空侵略者》(Space Invaders)式的空对地混战。眼见常规武器收效甚微,Wright 索性切换出终极大杀器,在剧烈的爆炸中直接将整座外星都市连同地表彻底轰平,只在行星周围留下一圈破碎的残骸光环。全场爆发出笑声与掌声,Wright 幽默地自嘲:“这下好了,我作为非暴力游戏开发者的名声彻底毁了。不过我们做了个漂亮的星环,这算是个不错的手工项目。” - 跨玩家星球探索完全基于后台异步下载,外星行为由系统根据原作者设置的性格参数(Personality Sliders)进行 AI 驱动。 - 首次接触包含类似《第三类接触》的音乐互动解谜,沟通结果直接决定和平建交还是爆发战争。 - 演示中摧毁外星文明的强力武器直接将地表结构解体,炸碎的物质在行星外围形成了如同土星环般的全新环带。 --- ### [55:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3321s) 科幻元游戏与“T 型”游戏架构哲学 在摧毁了外星城市后,Wright 将视角拉回太空,深入拆解了太空阶段的设计愿景。他计划在太空层级中融入一整套围绕经典科幻流派打造的“元游戏”(Meta-games): - **《世界大战》与监护者机制**:玩家可以扮演类似《地球停转之日》中的高级监护人与调停者。 - **物种飞升(Uplift)**:降临到只有原始生物的星球,放置类似《2001太空漫游》的黑色石碑,促使它们觉醒智慧,日后再返回查看其进化成果。 - **绑架与杂交**:自由调用生物编辑器,绑架外星物种进行杂交改造,甚至抓取生物去填充自己的“宇宙动物园星球”。 - **星际外交**:以《星际迷航》中的各种经典种族(如克林贡人、博格人)为标杆,通过极简的性格维度构建性格鲜明的外星文明,实现结盟、贸易或争霸。 Wright 指出,二十世纪三四十年代的经典科幻杂志封面通常由四种核心元素构成:载具、建筑、外星人与外星球——而《孢子》为其中的每一项都配备了深度编辑器。更重要的是,Wright 阐述了《孢子》颠覆传统的“T 型”游戏架构体系。 ```mermaid flowchart TD accTitle: 孢子 T 型游戏结构 accDescr: 展现从单向线性目标引导上升到顶层高度自由沙盒的 T 字形结构 subgraph SpaceTop["顶层:无尽沙盒与跨星系元游戏 (横向无边界展开)"] S1["物种飞升"] ~~~ S2["跨星系外交与贸易"] ~~~ S3["物种绑架与杂交"] ~~~ S4["星球浏览与摧毁"] end subgraph EvolutionVertical["纵向:目标导向的阶段式教学 (纵轴层层攀升)"] V4["文明阶段 (Civilization)"] V3["部落阶段 (Tribal)"] V2["生物阶段 (Creature)"] V1["潮池/细胞阶段 (Cell)"] V1 --> V2 --> V3 --> V4 --> SpaceTop end ``` 在大部分传统游戏中,沙盒模式往往被当作新手教程,玩家在沙盒里熟悉操作后,再去挑战具有明确目标的主线关卡。而《孢子》恰恰相反:它把具有明确目标的进化阶段(细胞、生物、部落、文明)放在前期,作为一套引导玩家掌握编辑器与底层动态模拟规律的庞大教程;一旦玩家历经进化冲顶进入太空阶段,游戏便彻底展开为一个水平无限延伸的终极沙盒。届时,玩家不仅能在不同恒星系之间进行横向穿梭,还能分形般地纵向潜入任何星球的文明、部落乃至生物阶段,自由漫游。正如 Wright 所言:“我们实际上是把更偏向目标导向的玩法放在了前面,作为日后终极沙盒的教程”。 - 太空阶段的灵感取自三四十年代科幻画作,涵盖飞升、绑架、第一次接触与深度外交等多种元游戏玩法。 - 外星种族行为模型以《星际迷航》种族为参照系,力求用最少的数学维度刻画出极端迥异的外星性格。 - 创新的“T 型”结构倒置了传统游戏逻辑:纵向的目标关卡是“教程”,用于培训编辑工具与模拟规则;横向延伸的太空沙盒才是真正的游戏本体。 ### [58:27](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3507s) 灵感起源:《十的次方》、德雷克方程与《Wired》杂志的秘密设计案 Will Wright 站在投影画面前,指着银幕上一个微小的方块解释道,刚才演示中体验的庞大区域,在整个宏观宇宙中其实仅仅占据了这么一个极其微不足道的角落。他坚信团队完全有能力在游戏中创造出一整座银河系,容纳成千上万、乃至数十万个可供玩家自由造访与互动的星球世界,让玩家在探索中走出专属于自己的游戏历程。谈及这一构想的源起,Wright 回顾了自己多年来对查尔斯和雷·伊姆斯夫妇(Charles and Ray Eames)经典纪录片《十的次方》(Powers of Ten)的迷恋;他一直非常崇拜伊姆斯夫妇展现尺度的视角,但长期以来苦于找不到一种方式,能将这种宏大尺度的逐层缩放无缝融入到一款自洽、好玩的游戏当中。 随后,Wright 转向了对 SETI(搜寻地外文明计划)与天体生物学的深入探索,并曾着手开发一款以此为背景的游戏原型。在此期间,他突然获得了一个顿悟:天文学中著名的德雷克方程(Drake Equation)——即通过星系恒星总数、拥有行星的比例、适居且孕育生命乃至智慧文明的概率来估算外星文明数量的经典公式——其公式内的各个因式参数,竟能够极为严整地一一映射到《十的次方》中不同量级的空间尺度上。于是,他决定将 SETI 概念与《十的次方》的尺度跨越合二为一,催生出了《孢子》(Spore)的核心架构。 ```mermaid flowchart LR accTitle: 游戏核心概念融合流程图 accDescr: 展示 Will Wright 如何将伊姆斯夫妇的《十的次方》尺度理念与德雷克方程的参数结合,最终诞生《孢子》的设计架构。 subgraph S1["灵感源流"] A["《十的次方》理念<br/>(Powers of Ten)"] B["SETI 与天体生物学<br/>(德雷克方程)"] end subgraph S2["架构融合与验证"] C["尺度级别与方程项<br/>精准映射"] D["设计案提前泄露<br/>(《Wired》杂志生命专刊)"] end subgraph S3["最终呈现"] E["《孢子》核心框架<br/>(从细菌到银河主宰)"] end A --> C B --> C C --> D D --> E ``` 大约在演讲前的一年,《Wired》(连线)杂志筹划制作一期以“生命”为主题的专刊,特邀 Wright 为其设计一个专属页面。Wright 当时询问对方是否可以完全由自己任意发挥,在得到对方肯定的答复后,他直接将这份尚未公开的游戏早期设计案原稿发了过去,甚至完全没向杂志社透露这究竟是什么。结果《Wired》在整整一年前就原封不动地将其刊登了出来,这意味着《孢子》最核心的设计文档实际上早已在光天化日之下公开发表了一整年,却无人发觉其真正的野心。 - **德雷克方程映射**:恒星数量、行星分布、生命诞生概率等天文学关键因子,直接转译为游戏中细胞、生物、文明与星际尺度的阶梯式跳跃。 - **《Wired》秘密发布**:在发布会一年前,《Wired》杂志的“生命”专题页上就已经白纸黑字印出了《孢子》的核心设计案。 ### [59:47](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3587s) 核心支柱:数据压缩、生成式系统与两百个原型 为了向大众传达这款概念庞杂的游戏,Wright 将整个玩家叙事压缩成了一句极其纯粹、直观的核心概念:“玩家将经历从细菌演变为银河之神的完整历程。” Wright 总结道:“and that we're kind of basically trying you know to replay you know the complete history and future of life”(我们基本上是在试图重现生命完整的历史与未来)。他打趣说,要把这么宏大的命题提炼成一句琅琅上口的宣发广告词实在太难,如果在场的同行谁能想出绝妙的宣传语,请务必私下告诉他。 Wright 指出,支撑起整个庞大体系的核心技术枢纽正是“数据压缩”。程序化与算法层面的极限压缩赋予了系统惊人的杠杆效应: 1. **赋能玩家创造力**:玩家不需要具备高深的 3D 建模知识,就能随心所欲捏造出独一无二的生物、载具与建筑。 2. **驱动生成式系统**:极度压缩的代码能够在运行时动态生成庞大而繁复的内容库,无需美术人员手工堆砌每一项资产。 3. **实现极致的内容轻量化与可移植性**:极为轻巧的数据体积让玩家创作的内容能够瞬间上传与跨网络流转,从而在每个玩家本地自动编织并呈现出一个近乎无限、无边界的鲜活宇宙。 为了攻克这一连串从未有人涉足的技术难关,开发团队把敏捷原型制作当作了一种坚定的信仰。Wright 透露,团队前后总共构建了 200 多个独立原型;事实上,过去几年间他在各种场合展示过的许多神秘实验性演示,实际上都是《孢子》不同模块的先导原型,只是当时没有对外公开它们之间的真正关联。 ### [01:00:54](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3654s) 突破想象力的瓶颈:向开拓者寄语与保密谢幕 回顾自己漫长的游戏开发生涯,Wright 坦言,以往开发《模拟城市》(SimCity)或《模拟人生》(The Sims)等作品时,最主要的阻碍通常是外界的阻力——比如必须花费巨大精力去向发行商、高管或团队证明这个奇怪的点子行得通,或者证明它能卖得出去。但在构思《孢子》时,情况却截然相反。Wright 直言:“when I look back on this idea you know the biggest obstacle I truly had was making myself believe that we could build the game it was my own imagination was the biggest bottleneck”(当我回首这个创意时,我真正面临的最大障碍,其实是让我自己相信我们能把这款游戏做出来;我自己的想象力才是最大的瓶颈)。他感慨道,一旦自己从内心深处彻底确信这款游戏在工程上是切实可行的,后续去说服工作室骨干成员以及 Electronic Arts 高管层的过程,反而变得顺理成章、毫不费力。 借此经历,Wright 向台下所有的游戏开发者与创意从业者送上了真挚的鼓励:如果在座的任何人脑海中曾经浮现过某些看似离经叛道、彻底游离于行业常规框架之外、“绝不可能成功”的怪异创意,切莫轻易将其弃置遗忘。他敦促大家隔一段时间就重新审视并唤醒这些尘封的灵感,因为谁也无法预料这些疯狂的种子最终会生根发芽、将大家引向何等壮阔的未知领域。 在演讲尾声,Wright 透露了一段花絮:本次在 GDC 2005 上的前瞻展示是他争得 EA 特别批准才得以成行的“绝密例外”。按照严格的保密协议,从走出这间演讲厅的这一刻起,直到同年五月的 E3 大展正式揭幕之前,他都不得在任何外部公开场合多谈半个字。在一片欢呼与喝彩声中,Wright 带着标志性的幽默对观众调侃道“你们可以继续鼓掌”,伴随着全场雷鸣般的掌声与起立致敬,这场奠定程序化生成游戏里程碑的划时代演讲圆满落幕。 ## Notable quotes > [02:04](https://www.youtube.com/watch?v=ofA6YWVTURU&t=124s) and so i contend that player stories will always be more powerful than scripted stories that we try to tell the players um — Will Wright > [01:00:54](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3654s) when I look back on this idea you know the biggest obstacle I truly had was making myself believe that we could build the game it was my own imagination was the biggest bottleneck — Will Wright > [55:21](https://www.youtube.com/watch?v=ofA6YWVTURU&t=3321s) 我们实际上是把更偏向目标导向的玩法放在了前面,作为日后终极沙盒的教程 — Will Wright ## Mentioned Will Wright, Chris, Spore, The Sims, The Sims 2, Pikmin, Katamari Damacy, Grand Theft Auto, Apple II, Myst, HyperCard, Rand Miller, Robyn Miller, Demo Scene, Pac-Man, Care Bears, Pokémon, Neopets, Ocean, Pixar, H.R. Giger, Dr. Seuss, American Greetings, Populous, SimCity, Legos, Star Wars, Risk, WarioWare, Inc., Diablo, Peter Molyneux, Sid Meier, Civilization, Game Boy, George Lucas, J.R.R. Tolkien, Robert Louis Stevenson, Mutual of Omaha's Wild Kingdom, Marlon Perkins, Kid Pix, SETI, 2001: A Space Odyssey, Star Trek, Charles and Ray Eames, Powers of Ten, Drake Equation, Wired, Electronic Arts, GDC 2005 ## Video description Original Video Description: Will Wright talking at the 2005 Game Developer's Conference about 'Spore', which looks like it could possibly be the best video game ever. https://www.gdcvault.com/play/1019981/The-Future-of-Content-(English Source Videos: https://archive.org/details/2005_GDC_-_Will_Wright_SPORE_Reveal_-_Camera_Angle_of_Wright_Only_-_2005-03-11_A https://archive.org/details/2005-gdc-will-wright-spore-demonstration-audience-reaction https://archive.org/details/2005-will-wright-keynote-spore-screen-only-2005-03-11 I tried to fix up this video as best as I could. Unfortunately the game sections were recorded off of a projector screen so it's pretty blown out. Original Videos (Dead Links): http://video.google.com/videoplay?docid=8372603330420559198 http://video.google.com/videoplay?docid=-775656482294094003 [Source](https://www.youtube.com/watch?v=ofA6YWVTURU)

Source quotes

[2:04] Will Wright: and so i contend that player stories will always be more powerful than scripted stories that we try to tell the players um [60:54] Will Wright: when I look back on this idea you know the biggest obstacle I truly had was making myself believe that we could build the game it was my own imagination was the biggest bottleneck [55:21] Will Wright: 我们实际上是把更偏向目标导向的玩法放在了前面,作为日后终极沙盒的教程

Source quotes

[5:24] Will Wright: I contend that player stories will always be more powerful than scripted stories that we try to tell the players. [39:48] Will Wright: I did not want to put the player in the role of, you know, Luke Skywalker or Frodo Baggins; instead I wanted to put the player in the role of George Lucas or J. R. R. Tolkien or Dr. Seuss, or whoever, you know? [41:49] Will Wright: More recently I've been thinking of story as really a side effect or a property of interesting experiences, not a prerequisite.

Original transcript

[0:00] all right well i you ready to go yep i'm [0:02] ready [0:04] hey everyone let's get this started um [0:07] we usually like to intro and thank our [0:09] speakers but you all know who will is so [0:11] if you can all just give will wright a [0:12] big [0:12] round of applause and thank you for [0:13] showing up so let's go [0:17] [Music] [0:19] um [0:26] yeah can you all hear me okay am i [0:29] amplified [0:30] oh good the voice of god um first of all [0:33] cell phones off [0:34] um no photography and fill out the [0:36] registration forms that i'll remind you [0:37] about later [0:38] um and they made me say all that stuff [0:40] so uh [0:41] let's see i wanna start introducing [0:44] chris right here who's going to help me [0:45] with my presentation a little bit today [0:46] she's been working with me as a [0:47] co-designer for about five years [0:52] and i'd like to start by also [0:56] apologizing first of all the translators [0:58] because i tend to go [0:58] really fast and because i always just [1:01] want to cram so much into an hour but uh [1:02] if you can't keep up feel free to just [1:04] make up stuff and [1:05] it might be better than what i'm saying [1:07] who knows um i also want to apologize [1:09] for the title of this talk in the [1:11] description [1:12] which i know is kind of hopelessly vague [1:14] um what usually happens every year is [1:16] gdc they call me up and they say we want [1:17] you to talk and [1:19] i say okay and then a week later to say [1:20] what you're talking about and so then i [1:22] think about it for a while and oh [1:23] i get a rough idea what i'm going to [1:24] talk about let me give them some kind of [1:26] generic title in that area [1:28] that's what usually happens but this [1:29] year was different this year when they [1:31] asked me what i was going to talk about [1:32] i just flat out lied um [1:35] but uh so i'm never going to talk about [1:38] the future of content um [1:40] i'm really more in fact going to talk [1:41] about a possible future of content you [1:43] know it seems like [1:44] the theme that i'm coming away from this [1:46] conference with so far [1:47] is that the next generation titles are [1:49] going to be incredibly expensive we're [1:50] going to have to build mountains of [1:51] content [1:52] and it's going to drive most of the [1:53] smaller players on the market [1:55] and i'd like to offer potentially an [1:57] alternate vision of that you know i [1:58] think that there is a way to deal with [2:00] that [2:00] more creatively [2:04] so maybe more accurate title for this [2:06] talk would be what i learned about [2:07] content from the sims [2:08] now you've probably heard me talk about [2:10] this a bit so we're going to do this a [2:11] little fast [2:12] um first of all content is very [2:14] expensive uh [2:15] this is a list of the content we had in [2:16] the sims 2. and [2:18] um compared to the sims 1 it's you know [2:19] just a mountain of work [2:21] to make that stuff um going back across [2:24] the games i've worked on [2:26] looking at the team scaling up i showed [2:28] you the slide last year if you saw my [2:29] talk [2:31] if you look at the content side of that [2:33] it's growing disproportionately to the [2:35] team size you know so it's scaling up [2:36] non-linearly [2:37] um so it's trending way up and i think [2:40] that's kind of what we're hearing at [2:41] this conference today [2:43] um you know one problem with that is [2:45] that you know for us to do twice as much [2:46] content [2:47] cost us about twice as much effort but [2:49] yet the value to the player [2:51] really isn't twice as much you know this [2:52] is just basic util [2:54] utility theory if you're on a desert [2:56] island and i give you an apple it's [2:57] worth a lot you know to keep you from [2:58] starving [2:59] once you have 100 apples you know the [3:01] 101st apple is not worth that much [3:03] um so i think you know we're fighting [3:05] kind of a losing battle here [3:06] um on the other hand we've seen that [3:08] players love making content um [3:10] time and time again every time we've [3:11] given the players opportunity to make [3:12] content they've just gone crazy with it [3:14] um on our website for the sims 2 we've [3:17] already got something like [3:18] 17 000 characters people have uploaded [3:20] 10 000 houses etc [3:22] um now obviously not all this content is [3:24] great but even if only 5 or 10 percent [3:26] of this is good in fact [3:27] the percentages are getting better i [3:28] mean in terms of what's really good [3:29] content players are making [3:31] it still represents a huge amount of [3:33] content that the players are willing to [3:34] build for us [3:35] this is one of our larger websites [3:37] phantom websites they just had a [3:39] celebration for their hundred thousand [3:40] object that people uploaded on there [3:44] another thing we found from the samsung [3:45] is that metres can be extremely useful [3:47] when we're doing games you know we can [3:48] learn so much from players by studying [3:51] uh [3:51] their play patterns what they buy what [3:53] they do the verbs they choose [3:55] um that we can use for various purposes [3:58] we see the players hate boundaries you [3:59] know we're seeing a trend toward more [4:00] open-ended games you know especially [4:02] unnatural boundaries where they just hit [4:03] kind of an invisible wall in the game [4:05] um one of the issues there is that as we [4:07] try to scale up the size of the [4:08] possibility space for the player [4:10] the development effort again is kind of [4:12] a non-linear thing to build an emergent [4:13] game it takes a lot of iteration [4:15] um we're currently trying to kind of [4:17] push that curve down a bit right now [4:19] by building more emerging gameplay and [4:21] we need to also start fighting it on [4:23] getting super content development um [4:26] i think there's kind of this inbuilt [4:28] assumption that the amount of value that [4:29] a player gets from content [4:31] is probably a function of the quantity [4:33] and the quality of it [4:35] um we can take it you know title like [4:36] pikmin and you know it doesn't have that [4:39] much content but it's very nice looking [4:40] and probably the area described [4:42] there roughly equals the player value um [4:44] we could take another game like katamari [4:45] damacy [4:46] you know which has a lot of content but [4:47] a much lower quality you know quality [4:49] and it feels like about the same amount [4:51] of value so from that i think we can [4:52] kind of infer [4:53] this gradient um that we're kind of [4:55] currently when we think about the amount [4:57] of content that we do in a game [4:58] you know we're following somewhere in [4:59] this gradient in terms of how much value [5:01] we're providing to the player [5:02] i think this is missing a very important [5:04] dimension though um [5:06] i think you know the ownership that we [5:08] can give players over content [5:10] uh feels like a very meaningful new [5:11] dimension that you know by pulling that [5:13] out just a little bit [5:14] we you know add tremendous amount of [5:16] area to that player value [5:17] you know so maybe we don't need as much [5:19] content and i didn't need it as high [5:20] quality [5:21] if the player made it they're really [5:23] into it [5:24] um this is my guy from grand theft auto [5:29] his name's mo he doesn't really do the [5:32] missions he's kind of a slacker [5:34] he rides his bike all day and learns new [5:35] little bike tricks on his bike um [5:37] but i love it when he goes into the cut [5:38] scenes you know this guy's wearing his [5:40] you know [5:41] cute underwear and his barbie watch and [5:42] they take him very seriously in the game [5:44] somehow [5:46] but you know i after i play the game i [5:48] go tell my my daughter and my wife's [5:49] stories about what moe did today [5:51] and you know i find that you know [5:53] ownership [5:54] over you know the experience translates [5:56] for me in a much more meaningful story i [5:58] never go tell somebody [5:59] you know the story about the cut scene i [6:00] just saw in a game but i'm always [6:02] telling people about the weird crazy [6:03] thing i did that was very you know [6:04] unique to me very you know uh [6:06] i really owned the experience and i was [6:07] driving the story i was playing the [6:08] story [6:09] um and so i contend that player stories [6:11] will always be more powerful [6:12] than scripted stories that we try to [6:14] tell the players um [6:16] and i'll kind of revisit that a little [6:18] bit later [6:19] so we've got these trends kind of going [6:21] from you know more linear gameplay to [6:23] more open-ended [6:24] um steeply rising content costs that [6:26] we're fighting [6:27] you know the value of player generated [6:29] content um metrics that we can capture [6:31] from the players [6:32] um so perhaps a more accurate title for [6:35] this talk today is going to be [6:37] what i learned about content from the [6:38] sims and why it's driven me to [6:40] procedural methods [6:42] now i started way back on the apple ii [6:44] uh [6:45] remember this a great machine [6:48] um and it was six times faster than a [6:51] teletype [6:54] this is actually from one of their ads [6:57] back then we programmed everything you [6:59] know just you know right down to the [7:00] metal [7:01] in you know machine code we actually had [7:03] memory maps we knew where everything was [7:05] in memory which is kind of cool [7:07] and actually you know what was in memory [7:08] about half of it was code and about half [7:10] of it was data [7:11] um you know the memory was in such short [7:13] supply that we had to you know [7:15] use as little data as absolutely [7:17] possible um today that ratio is probably [7:19] more like 101. [7:20] um then this happened uh cd-rom came out [7:25] and at that point you know compared to [7:26] the amount of memory we were dealing [7:27] with and these little tiny floppy disk [7:28] drives with 140k [7:30] on them and stuff um this thing was you [7:32] know a super tanker [7:33] we could not conceive of this much data [7:35] um you know where in the world are we [7:37] gonna get all the data to fill this [7:38] sucker up [7:39] um and so the trend started moving [7:42] toward [7:43] you know finding ways to create lots of [7:44] data to pile into our games you know and [7:46] one of the games that kind of really [7:47] exemplified this for me [7:48] well actually uh before i talk about [7:49] that i mean this basically was [7:52] kind of like the meteor that came in and [7:54] sounded the death knell for the [7:55] algorithm [7:56] you know in some sense uh this is when [7:58] data you know [7:59] started to really ascend and reign [8:00] supreme you know in our industry [8:04] one of the games that really exemplified [8:05] this for me was miss you know which was [8:07] a very elaborate [8:08] slideshow um it was actually done in [8:09] hypercard i mean robin and rand i [8:11] thought it was a brilliant game i loved [8:12] missed i loved all their games and the [8:13] fact that they managed to do this in [8:14] hypercars as a slideshow but yet still [8:16] making you immersed in this world was [8:18] absolutely remarkable to me [8:20] um and one thing that always occurred to [8:22] me though playing missed was that [8:24] you know there's this really cool [8:25] immersive experience but they had a [8:26] great time building this world [8:28] and i kept thinking that over and over [8:30] boy you know this is kind of fun to be [8:31] in [8:31] but i wish i could have made this world [8:33] that was the fun part [8:35] now you know while the rest of the [8:37] industry was kind of you know [8:39] going off into the data driven you know [8:40] let's fill up a cd-rom with lots and [8:42] lots of content [8:43] there was actually kind of another tribe [8:45] on the other side of the world that kind [8:46] of [8:47] kept the old ways [8:50] they basically worshipped the algorithm [8:53] kind of kept practicing [8:54] their black arts they adopted kind of [8:57] modern dress and technology [8:59] but uh and that tribe seems roughly [9:02] centered on [9:03] northern europe scandinavia ground zero [9:06] around helsinki [9:08] and we call it the demo scene [9:13] and these guys are basically crazy [9:19] they'll take you know these little tiny [9:20] things 64k [9:22] and through algorithm generate huge [9:25] amounts of [9:25] animation and visual stuff and music and [9:28] sound [9:29] and just amazing things algorithmically [9:31] procedurally [9:33] and just to give you kind of some sense [9:34] for how insane that is [9:36] if we take that 64k and we square it [9:38] square it again square it again [9:40] and again and again and again and again [9:43] okay that's a cd-rom over there okay [9:46] so you know why in the world were they [9:48] doing this what was the point [9:50] um they actually evolved out of the [9:52] cracking community they would crack [9:53] games and then uh [9:54] kind of sign their name put a little [9:55] screen at the front saying who cracked [9:57] the game because it was kind of a you [9:58] know [9:58] a little competition they had and then [10:00] eventually they got more animated more [10:01] elaborate visual [10:03] um eventually got very cool you know and [10:06] if you haven't seen any of these i [10:07] really encourage you [10:08] and then they have these rituals they [10:09] get together and practice their black [10:10] arts together [10:11] and um anyway i just kind of found this [10:14] just incredibly inspiring over the last [10:16] few years [10:18] so um probably a more accurate title of [10:22] this talk [10:24] is what i learned about content from the [10:25] sims why it's driven me to procedural [10:28] methods [10:29] and what i now plan to do with them okay [10:31] so the first thing i did was recruited [10:33] an elite team [10:34] of developers skilled in the old ways [10:40] um no slight tangent here sorry [10:45] my daughter was five you know many years [10:47] ago and i decided i was going to get a [10:48] really nice telescope and a really nice [10:50] microscope you know and [10:51] i could find a telescope everywhere you [10:52] know all these different science shops [10:53] nature company places like that [10:55] and i went to go look for a really good [10:56] microscope and i couldn't find one [10:57] anywhere and it really surprised me [10:59] because [11:00] you know i've used a really nice like [11:01] stereoscope in school and i really loved [11:03] what you could do with these things [11:05] and um this is just kind of you know [11:07] well's personal information you might [11:08] find valuable [11:09] uh really you know if i had to pick one [11:12] of the two i'd pick the microscope you [11:13] know especially with a kid these things [11:14] are so cool to play with [11:15] um and i'll tell you what you get [11:16] because now we have ebay and on ebay you [11:18] can get these things very easily whereas [11:19] before it was like a real treasure hunt [11:21] to find one [11:22] you basically definitely want one with [11:23] stereo and lighting overhead lighting [11:26] and they come in two types they come in [11:27] low power and high power and you want [11:29] the low power [11:30] as counterintuitive as that sounds i [11:31] mean and just you'll put your hand under [11:32] this thing and you'll spend an hour just [11:34] obsessing over your hand [11:35] and it's like a religious experience but [11:38] anyway back to our story [11:39] um so [11:43] so we recruited this elite team of [11:45] developers um [11:46] then i put them in a state-of-the-art [11:48] hidden development facility [11:50] um and started working on this thing a [11:53] little while back [11:54] um now what i'm going to show you today [11:57] uh actually is a little bit of the [11:58] results of what we've been working on um [12:00] normally i don't really do demos but [12:02] actually i really wanted to [12:03] demo this first to this community just [12:05] because i feel so much affinity for this [12:06] community rather than like an e3 [12:08] marketing type thing [12:10] um so what i'm going to show you first [12:11] of all it's very unfinished [12:13] it's kind of held together with you know [12:14] bailing wire and duct tape um [12:17] a long way to go but i think you know i [12:18] can show you enough of it to where [12:19] you'll get some sense [12:21] of why i'm excited by procedural methods [12:23] and i think it is relevant to kind of [12:24] you know what people have been seeing at [12:25] the conference here today um [12:28] we're calling it spore and that's we're [12:31] going to move over to this screen here [12:38] okay we start [12:41] here and we're in the land of the very [12:43] small [12:46] i'm actually controlling this little guy [12:47] around here um this is basically [12:49] cellular scale we have like some fluid [12:51] dynamics [12:52] um this basically is a very analog [12:55] fluid version of pac-man so basically [12:58] i'm [12:58] trying to eat these little dots uh these [13:00] little bad things chase me [13:02] um certain ones like these guys can kill [13:04] me too [13:05] ah that's bad um i have certain [13:08] abilities in this world [13:09] but uh this world is pretty much [13:10] entirely procedural [13:12] it's usually generated now the abilities [13:14] i get from this guy [13:15] after i eat enough little of these food [13:17] pellets i get to lay an egg that little [13:20] yellow thing is my egg here [13:21] um i now actually get to go to the next [13:23] generation of the guy i can click on the [13:24] egg [13:26] and go into an editor if we can bring [13:28] the lights down just a little bit if [13:29] that's possible [13:30] um so here's a little editor i have very [13:33] simple [13:34] parts that i can kind of build for this [13:35] guy i have mouthpieces that influence [13:37] what i can eat i can either eat filter [13:38] food or i have a probiscus i can poke [13:40] into things [13:41] um i have different movement methods [13:43] flagella cilia little jet [13:45] um and weapons like name assists spikes [13:47] whatever so i'm gonna take a little [13:48] spike here [13:49] put it on the front of the guy um and [13:51] then go back into the game [13:53] and i've actually modified this guy's [13:54] abilities now so whereas those brown [13:57] things used to kill him [13:58] now he can go in and attack them [14:02] and i still can't go after the brown [14:03] guys but um now i have a new food source [14:06] now i could change [14:07] my movement and get faster i could get [14:09] stronger um so basically the character [14:11] that i'm controlling [14:12] here is something that i've designed in [14:13] the editor um [14:15] now over time as i'm playing this game [14:17] uh the camera kind of pulls out the [14:19] depth of field [14:19] you know kind of res is down this guy [14:22] grows as it gets more complex [14:24] um i'm going to skip ahead just a little [14:26] bit in his life cycle what you're [14:27] actually doing is you're playing through [14:29] several generations here [14:30] um and at every generation you get the [14:33] ability to go in and kind of modify it [14:35] so this is a bit later in the game we've [14:37] actually gone now from a 2d [14:39] to a 3d environment and now rather than [14:42] living in a little you know drop of [14:43] water [14:44] we're living kind of under the ocean now [14:47] this entire world you see here is [14:50] procedurally generated [14:51] um especially the guy which i've created [14:54] as a player when i say procedural i also [14:56] kind of include player [14:58] creating things being very malleable in [15:00] a procedural way so for instance um the [15:02] player creates this character but the [15:03] animations [15:04] behaviors textures are all generated [15:07] procedurally [15:08] now this guy has a life cycle very much [15:11] like the little cell guy did [15:13] he also competes with other creatures in [15:14] the environment [15:16] another thing to realize here that's [15:17] kind of different is that all the [15:18] creatures in this environment [15:21] are coming from other players look [15:24] around here a little bit [15:26] let's run over those guys [15:33] now when i say coming from other players [15:34] what i mean is that they are [15:36] asynchronously [15:37] being sent to your computer to fill out [15:39] your world we're actually building a [15:40] balanced ecosystem depending on what [15:42] your world needs [15:43] we deliver other players content to pull [15:44] out that world around you [15:46] now like the cell game you know this guy [15:48] when he reproduces he lays an egg [15:52] click on the egg and now we go into the [15:54] editor [15:58] now this is the editor for this guy um [16:00] you can see he's kind of made out of [16:02] parts [16:02] um the editor here is roughly a mixture [16:04] of kind of mr potato head [16:06] erector set in clay i can grab the parts [16:09] and manipulate them [16:10] um i could even pull them off um [16:14] and worf them so i'll pull that off i [16:16] can grab the [16:17] spine here manipulate it you see there's [16:19] an underlying skeleton here which gives [16:21] me some clue is kind of internal [16:22] structure of the guy [16:25] and like clay i can roll the mouse wheel [16:27] and actually sculpt him smoothly [16:29] to give him his outer form [16:34] now these things are built out of [16:36] functional parts that actually have [16:37] meaning to the guy [16:38] um in terms of how he's going to behave [16:40] and live in this world [16:42] these parts also have are mutable [16:48] so let's pull off this tail here [16:52] the part categories um come in [16:53] functional divisions [16:55] so these are graspers um a lot of the [16:57] parts also [16:58] have some functional value besides their [17:00] primary one so that's a grasper as well [17:02] as having a little bit of a weapon [17:03] ability [17:04] but it's primarily grasper this is a [17:08] three-jointed leg [17:08] make a little bit smaller [17:12] um these are feet so basically i've [17:15] taken my little swimming guy i can also [17:16] scale the parts as i put them on [17:18] um so i've taken my little swimmy guy [17:21] and basically pulled his fins off which [17:23] is going to mean he's not going to swim [17:24] too well now [17:25] um i'll put let's see [17:28] another leg there [17:32] so this guy is going to be a tripod [17:35] guy um so the player can build like [17:38] really unusual characters in here [17:39] the system will analyze what they built [17:41] and then kind of bring it to life all [17:43] procedurally [17:44] so um we'll bring this guy back into the [17:46] game [17:48] and because he lost his fins he doesn't [17:51] really swim too well now [17:52] so i'm gonna actually have him kind of [17:54] leave the ocean here [18:05] now the walking you see here is all [18:07] procedurally generated based upon an [18:08] analysis [18:09] of the skeleton that you created um in [18:12] fact every verb in the game [18:13] is as well and once again you know the [18:15] world you're seeing here is procedurally [18:17] generated [18:17] other creatures coming from other [18:19] players um [18:21] like the other levels this guy has to [18:22] kind of survive this guy's a carnivore [18:24] so i'm going to have him [18:24] eat and have him come in and attack this [18:27] guy [18:36] now the way he's fighting is based upon [18:39] where i put [18:40] his weapon which is on his tail [18:48] so you can see you know i'm not close [18:50] enough it won't hit [18:52] so i basically have to chase luckily i'm [18:53] faster than this guy oh there we go [18:55] get off my head yeah okay [18:58] so we killed this little guy [19:04] now we can actually start eating him [19:13] now one of the cool things about [19:14] procedural verbs is that they can be [19:16] cleanly mixed [19:17] um for instance i can tell this guy take [19:19] a bite out of this guy [19:20] and then tell him to walk at the same [19:22] time and use that to actually invent a [19:24] new verb [19:25] which is to drag [19:32] one of the real wins though is that [19:33] whatever i've created in this game is [19:35] going to behave differently i mean you [19:36] know it's just going to be fundamentally [19:37] different behavior than anything that [19:39] you might create in the game [19:40] uh which is a huge win [19:44] and oh okay he's a bad guy [19:52] this is where you say run away [19:56] now your game actually builds kind of a [19:58] balanced uh ecosystem so you're not [20:00] always the top of the food chain here [20:02] um on the other hand this guy's kind of [20:04] fast [20:05] and you know maybe he has something on [20:06] his stealth and maybe i can outrun these [20:07] stupid little hoppy guys [20:09] and he'll go eat one of those guys [20:10] instead [20:14] let's see [20:17] i think i might be safe [20:21] yeah okay so one of the next things you [20:24] do [20:24] basically you live out lifespan of one [20:26] of these creatures um [20:28] and then you basically mate which earns [20:30] you access to the editor and you earn [20:31] currency at this level the game so i can [20:33] do a mating call [20:42] and try to find somebody to mate with [20:43] okay we'll see how that goes [20:52] i'll kind of approach carefully [20:59] yeah that sounds good okay [21:02] okay and then we get to the procedural [21:10] [Applause] [21:14] mating [21:22] [Applause] [21:28] okay enough of that silliness [21:31] then we lay an egg and that gives us [21:33] access to the editor where we get to [21:34] create the next generation of creature [21:36] um now i want to give you a little bit [21:38] of a sense for [21:40] how open-ended this really is i'm going [21:41] to load a few creatures that we've [21:43] created in the editor [21:44] um just give you a sense for how [21:45] creative the player can be with this [21:47] um this guy [21:51] i call him buttface but notice he has a [21:54] rather non-standard leg configuration [21:57] but um [22:01] but still it moves very smoothly um very [22:03] realistic very lifelike [22:04] um you know one little kind of subtle [22:06] trick we're playing with here is that uh [22:08] you know people have never seen buttface [22:10] before so they don't know how he's [22:10] supposed to move [22:12] as opposed to humans [22:15] here's another one now all the ones i'm [22:16] showing you here are actually running [22:17] the exact same script [22:19] um this guy's got kind of a large head i [22:21] call him tweety bird [22:23] it's kind of like an suv or this is the [22:26] ford explorer like whoa [22:30] [Applause] [22:31] [Music] [22:35] i suspect he would have low fitness in [22:37] the world [22:41] let's see this guy called hang dog he's [22:43] very cool [22:45] he kind of walks cool too [22:49] but again you know basically the same [22:51] code running on all these things that [22:53] you know analyze what the player's done [22:54] and then kind of animates them brings [22:56] them life there's also a whole [22:57] procedural texturing step that we don't [22:59] have integrated into this [23:00] like i can't show you right now but the [23:01] ones in the world you saw that were [23:02] procedurally textured [23:04] this guy i call dilemma [23:10] so the function of these guys is a you [23:12] know a combination of the parts that you [23:13] choose for them [23:15] and also kind of the topological um [23:18] structure that you set up so this guy [23:20] just happens to have a very fast walking [23:21] gate [23:22] and it's a property of kind of the way [23:23] the procedural animation deals with that [23:25] skeleton [23:26] so a lot of the kind of performance of [23:27] these guys is going to be emergent you [23:28] know [23:29] from the system and of course i insisted [23:32] to the team that we had to do a care [23:33] bear [23:34] so here's our care bear [23:41] i'm gonna make him a really mean [23:42] carnivore you know [23:47] so anyway that's our little evolution [23:48] game so far [24:01] so one of the things we want to do with [24:03] these with this editor here is to make [24:05] it a toy it's in line with the gameplay [24:07] when people were playing the sims they [24:08] had to go out to a separate you know [24:09] place download a tool [24:10] do something with it then upload it to a [24:12] website we really i want to burn the [24:14] tools into the gameplay [24:16] and then make the portability of the [24:18] content transparent to the player [24:20] i don't have to go to a website and say [24:21] i want some cool you know creatures from [24:22] my planet [24:23] they should come to me automatically it [24:24] should be able to analyze the way i'm [24:25] playing and bring them to me [24:26] automatically [24:28] um i've seen a lot of evolution games [24:30] and most of them really did not you know [24:32] i didn't feel like i owned the creatures [24:33] that were created because it was a very [24:35] simple like you know pick from the set [24:36] of parts [24:37] or these three slots you know a head [24:38] body tail to kind of get that [24:41] um as a kid you know i grew up you know [24:43] just loving dinosaurs you know and [24:45] kind of the generation i'm from that was [24:46] like what kids did and you know [24:48] even then it was very much a paper rock [24:49] scissors thing you know by learning the [24:51] different types and the names of the [24:52] dinosaurs you know it's kind of cool [24:53] that i knew what an ankle source was my [24:54] mother didn't [24:55] you know and that gave me a certain [24:56] ownership of those um you know today [24:58] it's more likely [24:59] something like pokemon you know these [25:01] things you know have very elaborate rule [25:02] sets and structures that kids learn [25:04] and then you know they feel some sense [25:06] of mastery over these things [25:08] um now just imagine if they could create [25:09] them you know from scratch [25:11] and that's kind of what we're trying to [25:12] do um neopets as well you know one thing [25:14] that actually [25:15] is amazing is that as you start giving [25:16] these things properties you start [25:18] naturally mapping narrative into them [25:20] um going back to care bears for a second [25:22] uh they were i don't know how many of [25:23] you know but they were originally just [25:24] greeting cards [25:25] they were invented by american greetings [25:27] for greeting cards if you're feeling bad [25:28] you give this [25:29] care bear card or if you're feeling you [25:30] know this way you give that care bear [25:31] card [25:32] um they had very specific properties and [25:35] specific friends and little superpowers [25:37] and stuff [25:37] they're actually web apps you know right [25:38] now i filled out one of the little [25:40] psychology apps [25:41] that tells you which care bear you are [25:42] turns out i'm sunshine bear on the [25:44] bottom corner there [25:46] um and just when you think you've got [25:48] all the care bears memorized they go and [25:49] invent care bear cousins [25:51] which they're not bears at all they're [25:53] all these weird things [25:54] except for that little yellow guy at the [25:55] bottom looks like a bear with a tail and [25:57] they don't even live in the same place [25:58] they live in forested feelings the care [25:59] bears live in care a lot [26:01] um so the care bears are up here in [26:03] caroline and apparently forced to [26:04] feelings is below carolina because when [26:06] they fall off the cloud they land there [26:08] and earth is even below that because [26:11] when they have to go on missions they go [26:12] all the way down to earth which would [26:13] put them somewhere in like low earth [26:14] orbit [26:15] um and they collectively refer to this [26:17] as the kingdom of caring [26:19] um and they never specifically say it [26:21] but they kind of imply that earth is the [26:22] kingdom of who gives a [26:24] because that's that's where they always [26:26] have to go on their missions [26:27] but [26:31] but i digress [26:35] our lead artist ocean quickly coined [26:37] this phrase artist in a box you know [26:38] we're looking at all the things that an [26:40] artist would typically have to do to [26:41] create content in a game and figure out [26:43] you know how can we possibly [26:44] automate all these things and then give [26:47] you know a very simple set of controls [26:49] to the player to where they feel this [26:51] incredible creative leverage [26:53] um the stylistic dimension that we would [26:55] like the players to kind of traverse [26:57] is somewhere between like pixar and [27:00] giger [27:01] i don't think we're going to get quite [27:02] that far but i think it's the right [27:04] dimension so we want players to be able [27:05] to do like these [27:06] kind of dr seuss worlds or these very [27:08] realistic kind of scary worlds [27:09] in the game um in essence what's [27:12] happening is we have the player here [27:13] manipulating a very simple little system [27:15] um and that's their leverage on the [27:17] thing [27:18] the computer takes that very simple [27:20] little system that they've built [27:22] um and then generates out the much more [27:24] elaborate data sets the mesh [27:26] the textures um this is actually some of [27:28] the procedural texturing that we're [27:29] doing where the player has [27:30] control over which scripts run and they [27:32] can be combined to get a lot of really [27:34] interesting results and [27:35] they actually analyze the topology of [27:36] the mesh so you get it knows where the [27:38] back is and the belly and all the stuff [27:41] also the animations which you saw and [27:44] then eventually the behavior of these [27:45] things [27:46] which can be player driven as i was [27:48] driving my little character around the [27:49] world [27:50] or they can be npc driven you know when [27:52] i pull in other players characters and [27:53] they're living in my world [27:54] um you know the computer is now [27:56] controlling those things [27:58] eventually you get this you know very [27:59] elaborate cool looking realistic thing [28:01] in the world and what this represents [28:03] actually is a creative amplification [28:05] of the player's efforts the computer is [28:07] really an amplifier here a creative [28:09] amplifier on what the player's done [28:11] and if we look at how small we can [28:13] actually compress these representations [28:15] of you know the stuff that i was doing [28:16] in the creature editor could probably be [28:18] compressed down to about 1k [28:19] bytes in memory with animations and [28:21] textures and all that stuff [28:22] it's more like five megabytes and so [28:24] that's something like a five thousand to [28:25] one compression [28:26] now that buys us a lot of things first [28:29] of all it gives the player the [28:30] tremendous leverage creatively [28:33] it means that we can store a huge number [28:35] of these things in a very small database [28:37] so even if you don't have a network [28:38] connection we can give you a little [28:39] database [28:40] you know on the cd with your game that [28:42] contains tens of thousands of these [28:44] things already if you're not connected [28:45] to the net [28:46] it also gives us very low friction for [28:48] moving these things they're very cheap [28:50] to move around [28:51] from player to player peer-to-peer or [28:52] server host [28:54] and also the same things that we do to [28:57] give the player leverage are almost [28:59] exact same things you would want to do [29:00] to build a generative system [29:02] um the computer can be playing with [29:03] those knobs just as well as the player [29:05] um very much like you know when [29:07] somebody's making a face or character in [29:08] the sims [29:09] they're doing sliders it's going through [29:10] some process creating the finished [29:12] content [29:13] and this is kind of basically mapping to [29:15] the idea of a genotype going through [29:16] embryology [29:17] to an initial phenotype and that [29:19] compression is vital for this whole [29:21] thing [29:22] and those levers again as i said can be [29:24] pushed either by the player or by some [29:26] generative system [29:27] so the pollination that's happening here [29:29] is as you build a creature [29:30] um it actually gets sent up to our [29:33] server this is one model [29:34] actually a lot of models could work here [29:35] could be appeared up here as well um [29:37] where we're sorting the best content and [29:38] we're categorizing it they're also [29:39] categorized [29:40] by things like their life cycle and how [29:42] they behave in the world how big they [29:43] are what they eat [29:44] um so we can do parametric mapping to [29:46] the content [29:47] um in the meantime your world kind of [29:50] requests we need people to fill out the [29:51] world that i'm going to be competing [29:52] with [29:53] and they get pulled in to build a little [29:54] ecosystem around you um [29:57] so that's basically kind of the [29:58] pollination idea here where the world is [30:00] being built around you continuously as [30:01] you play [30:03] um let me go back over here [30:08] to the editor now there's one more [30:11] palette at the bottom here [30:13] which is the brain palette now this is [30:16] what you're really trying to do here [30:17] is you're trying to save away a little [30:19] bit of funds as you're playing the game [30:21] and slowly invest them in the brain this [30:22] is kind of like your say your college [30:24] savings account [30:25] you know a lot of times you're buying [30:26] stuff oh i'm getting beat up i'm going [30:27] to buy you better weapons or i'm going [30:29] to hide or run faster [30:30] but if you can save a little bit extra [30:31] you start investing in the brain and [30:32] work your way up kind of this ladder [30:34] now once you get the highest level brain [30:36] your creatures actually become sentient [30:38] and you move to the next phase [30:42] now at this level the game actually [30:44] mutates from kind of a first person [30:46] eater [30:47] to an rts now we're controlling a whole [30:50] little herd of these guys [30:52] um now one thing that uh [30:56] i talked about landmarks last time in my [30:58] talk last year and i'm going to show you [30:59] just a few of the landmarks for me in [31:01] the design [31:02] one of the landmarks is i've always been [31:03] fascinated with the idea of monkeys with [31:04] guns [31:06] and i don't know why you know but just [31:07] what happens when you give monkeys guns [31:09] you know it's just [31:10] there's something cool about that so at [31:12] this level rather than you know [31:14] the creature is now set you've evolved [31:16] the creature um at this level i can buy [31:18] them tools [31:19] um and i in this case i'm shopping for [31:21] tools and i'll talk about the shopping a [31:22] little bit later in this case i'll buy [31:23] them a spear rack [31:26] they get excited we're seeing kind of [31:28] more elaborate procedural verbs at this [31:30] point you know they'll kind of have a [31:31] response [31:33] decide what they want to do with it [31:40] now before they get into trouble i'm [31:41] going to buy them a campfire here um [31:46] these guys are still a bit excitable uh [31:51] but i'm actually dealing with the social [31:52] dynamics at this point um [31:54] depending on how they play them they can [31:55] become very emotional or very logical at [31:57] this level of the game [31:58] uh let me buy them a drum [32:06] i bought them at drama now these guys [32:07] decided to start playing the drum [32:09] um and [32:12] this is actually procedural dance you [32:15] know based upon [32:16] my creature [32:17] [Applause] [32:25] so at this level of the game you're [32:26] playing it kind of like populist or [32:27] something i'm buying them tools it's [32:28] influencing what they can do i can [32:29] control groups of them now i'll send [32:31] them out on like hunting parties [32:33] and i'm in fact competing with other [32:34] tribes um [32:36] at this level the latter that i'm [32:37] climbing is i'm really upgrading their [32:39] hut [32:40] um and so i'm gonna buy them a better [32:42] hut here which will allow me to buy the [32:44] better tools [32:46] um and also allow my my tribe to grow [32:48] and i'm competing with other tribes for [32:49] members [32:50] uh they have to do with food and stuff [32:51] like that um [32:54] now they're just basically living in [32:55] this little corner of the world i'm [32:56] going to skip ahead a bit here [32:57] um if as i upgrade the hut here [33:00] eventually i [33:01] upgrade it uh to the next level which [33:03] brings me um [33:04] to get the next game which is the city [33:08] game [33:10] now at this level we're basically [33:12] playing a very very simple version of [33:13] simcity [33:14] um i can you know decide what buildings [33:17] to make them i [33:18] build roads uh they have needs [33:21] um and basically i'm trying to keep them [33:23] happy in their little city you know and [33:24] again i have [33:25] complete creative control over [33:28] everything you see here [33:29] you know so i am you know the player is [33:30] the author of this little world [33:38] now like the tribal mode um we have a [33:40] shopping thing here now [33:41] the shopping is populated actually again [33:44] with other players content [33:46] um so if i don't feel like building [33:48] something you know and i just i see [33:49] something really cool that another [33:50] players made it'll try to learn the [33:51] style [33:52] that i'm using and try to populate my [33:55] shopping catalog [33:56] with things in that style and then i can [33:58] pick them [33:59] here and then buy it [34:03] or you know if i see something i want to [34:05] modify or i want to build something from [34:07] scratch [34:07] i can actually go into the building [34:08] editor um which [34:11] and we're trying to make the you know [34:12] all the editors in the game um [34:14] feel very very similar uh this is using [34:17] similar kind of morph block stuff [34:19] you know basically everything in here [34:20] has like a handle and i can kind of play [34:22] around with it [34:23] shape it you know like these little um [34:25] we have categories you know from very [34:27] platonic solids [34:28] that have certain handles on them [34:36] now the building editor is a little bit [34:38] closer to legos really [34:40] um i can grab an object and i can have [34:42] it intersect [34:44] or stack um you know try to find the [34:46] simplest possible operators [34:48] you know the average person can very you [34:49] know quickly learn and this becomes a [34:50] toy to them you know they really have a [34:51] lot of time and we found this a lot in [34:53] the sims build mode by the way [34:54] the players would go into that mode [34:56] there was a little bit of a learning [34:57] curve but they had kind of immediate you [34:58] know [34:59] feedback and they got so much creative [35:01] um [35:02] reward for doing that that uh i've been [35:05] looking at how we fold that experience [35:06] more and more of the gameplay okay so we [35:07] basically added this little mushroom [35:09] thing to our building we'll go back in [35:11] the game [35:13] there it is over here so they're they're [35:15] happy they like it [35:19] the primary feedback for me in this game [35:21] is kind of the [35:22] happiness of my people and they you know [35:23] now we're dealing with like large-scale [35:25] kind of social interactions between the [35:27] characters um [35:29] as i mentioned before the tribal game [35:30] you know uh this is my city but i'm [35:32] actually competing you know with other [35:33] cities um if i scroll across the map i [35:35] can look for neighbors [35:37] so here's another city now this city was [35:40] created [35:40] in the exact same editor um just using [35:43] you know slightly different primitives [35:44] um different texturing you know from the [35:46] procedural texturing but just as [35:48] to give you a sense of the variety of [35:50] style that a player can do so our city [35:52] was kind of what we call maybe dr seuss [35:54] style [35:55] um this is more kind of [35:56] techno-industrial star wars style [36:02] but these kind of demonstrate a little [36:03] bit the kind of pixar pixar to gear [36:06] dimensions now these guys also will have [36:07] a very different culture than my guys [36:08] you know depending on how they [36:10] we evolved and how i you know dealt with [36:11] him or how the [36:13] computer dealt with these guys uh during [36:15] evolution and tribal [36:16] these guys are really kind of more [36:17] aggressive a little bit more industrial [36:20] um at some point once you upgrade your [36:22] city enough um [36:23] you get the ability to start building [36:24] vehicles [36:27] this is how the cities actually interact [36:29] as a tribal you're actually sending [36:30] people from tribe to tribe and doing [36:32] raids [36:33] or you know trading having you know you [36:34] can have kind of good or bad [36:36] relationships with the neighboring [36:37] tribes [36:38] at the city level we basically have kind [36:40] of three ways to [36:41] capture or ally with another city [36:43] they're basically military economic or [36:45] cultural [36:46] um depending on kind of the cultural [36:48] flavor of my people [36:49] um so they don't look very friendly and [36:51] they're heading our way [36:53] so we can open our vehicle editor which [36:56] works very much like the building editor [36:57] there's a shopping alternative populated [36:59] with you know items that other players [37:01] have created [37:02] um and i'm just going to buy something [37:04] here [37:05] like that we have parts categories these [37:07] have functional meaning so propellers [37:08] you know can help the thing fly [37:10] uh certain things enable it to go into [37:11] water or go on land but um i'm just [37:13] gonna go ahead and go for this goofy [37:15] looking thing this is kind of the [37:17] more of a nazca looking and there they [37:19] go [37:21] god knows where they're a little goofy [37:24] um [37:24] yeah go to that city go go go no not [37:27] that way [37:29] okay so they're peaceful cowards so it's [37:32] hopeless [37:33] i don't know in the meantime these guys [37:35] are [37:37] you know not looking real friendly [37:43] so this this is kind of militaristic but [37:45] uh there are [37:46] more peaceful avenues to kind of [37:48] capturing cities [37:50] at this point in the game we're actually [37:53] kind of entering the uh [37:54] what i call the civilization game um [37:57] and this is a very simple version of a [38:00] game like sieve or risk [38:02] as we pull back now for the first time [38:07] you can see the entire world other [38:08] cities now one little trick you might [38:10] notice here [38:11] is that as i pull away from a city it [38:14] actually gets very exaggerated [38:15] um this is both for aesthetic and [38:17] functional reasons like [38:18] as i pull into the city notice how it [38:20] pops to a realistic scale but as i pull [38:22] out [38:22] it actually becomes an icon um that [38:24] indicates actually the level of [38:26] micromanagement [38:27] um at this level the city is one big [38:29] button and it has a very simple menu if [38:30] i want to go in and actually go and [38:31] still modify the cities [38:32] i'm still building paris while they're [38:33] attacking it i can still zoom in and [38:35] click the individual buildings and bring [38:36] them into the editor and stuff [38:38] um so basically you know you play this [38:40] game the civ you know game [38:42] you know and slowly conquer the world [38:44] you know city by city kind of my [38:45] continent until pretty much [38:47] you have owned this entire little world [38:49] and built it on the way [38:51] so that's that [39:00] now [39:06] now the next let me just [39:10] break that down a little bit now [39:13] in this part of the game we're dealing [39:15] with a different type of pollination as [39:16] i mentioned um we have a buy mode [39:18] in a lot of these levels and here we're [39:20] taking other players content [39:22] trying to map it parametrically to your [39:23] play style to populate your buy mode [39:25] um then you know you might choose to buy [39:28] one or actually build your own when you [39:29] buy something though [39:30] um we're actually using that as a [39:33] positive feedback [39:34] on what was in the database so we're [39:36] continually resorting the database on [39:37] the top end [39:38] trying to discover the best feedback [39:40] based upon how many players actually [39:42] used it [39:43] and so there's continual sorting of all [39:45] the content that we're accruing in the [39:46] master database [39:48] uh at this point you can see this game [39:50] actually has some similarities to wario [39:52] incorporated [39:53] um in some sense i wanted this to be an [39:55] homage to all of my favorite strategy [39:57] games [39:58] and i know it's kind of like one of the [39:59] rules of thumb in the game is don't mix [40:00] genres [40:01] and so i've always wanted to break that [40:02] rule really badly [40:06] so you know we basically have kind of [40:08] the tide pool level which feels very [40:09] much like pac-man that was kind of what [40:11] you know evolution which is like diablo [40:13] counter-intuitively and i say like [40:14] diablo because it's kind of like you go [40:15] into this world [40:16] and you you know hack and slay monsters [40:18] and if you make it out alive you get to [40:20] go back to the village and buy better [40:21] armor better sword [40:22] and then go out back and fight you know [40:24] better monsters and that's kind of the [40:26] dynamic that you're going through when [40:27] you're going to the editor [40:28] living through a generation and then [40:29] going back um [40:31] the tribal level you know really feels [40:33] to me like population you know i wanted [40:34] to be kind of an homage to populists [40:36] which i loved um simcity [40:39] of course i had to uh and then sue [40:43] which i always loved you know and i [40:44] wanted the top but i wanted it to be a [40:46] very [40:47] light game experience compared to what [40:49] most of these games have turned into [40:50] i wanted it to be about 20 percent of [40:52] the best of simcity [40:53] or 20 of the best of uh civilization [40:57] because i as a hardcore gamer just don't [40:59] have the time to invest in these games [41:00] nowadays i find myself spending a lot [41:02] more time playing my game boy [41:04] i'd much rather have a lighter [41:05] experience that distills all the core [41:07] elements out and so [41:08] i'm fully expecting about 40 of this [41:10] experience to be really aesthetic [41:12] that the player spends a lot of time [41:13] just kind of building a creative world [41:14] and for them [41:15] it's part kind of light gameplay but [41:17] still very open-ended you know very [41:19] interesting strategies [41:20] but very understandable simple systems [41:21] that anybody can kind of pick up and [41:23] play like they can play a gameboy game [41:24] um for every level of course we have an [41:26] associated editor that becomes kind of [41:28] the main thing you're interacting with [41:29] you know at that level um and really [41:33] you know i did not want to put the [41:34] player in the role of you know luke [41:36] skywalker [41:37] or frodo baggins instead i wanted to put [41:39] the player in the role of george lucas [41:41] or jr tolkien or dr seuss or whoever you [41:44] know i wanted the player to feel [41:45] like you know they had built an entire [41:47] world you know going through these [41:48] different levels [41:49] um let me talk just a little bit about [41:51] the story uh just to deflect the issue [41:54] um because i don't have enough time [41:55] today to really go in depth of it [41:56] but more recently i've been thinking of [41:58] story as really a side effect or a [42:01] property [42:02] of interesting experiences not a [42:04] prerequisite [42:05] you know so many games that had no story [42:07] where i went in and had a great story [42:08] you know i don't even know what the [42:09] story of grand theft auto was but my [42:11] story was really fun [42:12] i had a great time you know and it was [42:14] very you know relevant to me i own that [42:15] story [42:16] um and i know it wasn't you know star [42:19] wars or hamlet or anything but [42:21] still it was my story um you know i [42:23] think we use the stories as a filtering [42:24] process [42:25] on our experience um they emphasize the [42:28] causality [42:29] of the situation if you look at a movie [42:30] the director has the opportunity to show [42:32] you exactly [42:33] why this guy got upset and kicked this [42:35] and broke his foot you know [42:36] and we don't necessarily have that level [42:38] of control you know [42:40] from our side but the player does the [42:41] player is always there having the [42:43] experience happening to them [42:44] um also stories are very useful for [42:47] portability you know communication it's [42:48] way for us to take an experience that [42:50] we've had some other place some over the [42:51] point in time [42:52] conveying somebody else and they're also [42:54] in some sense compression and [42:55] abstraction [42:56] where we can get general kind of lessons [42:58] that we might apply to other situations [42:59] even though the stories don't exactly [43:01] match [43:01] we can abstract up the story to kind of [43:03] a schema learning [43:05] um and i mentioned this actually uh [43:08] last year a little bit robert louis [43:10] stevenson when he did treasure island [43:11] the first thing he did was draw a map of [43:12] the island [43:13] and then his eye just roamed that map [43:15] kind of imagining all the things that [43:16] could play out [43:17] and in some sense the story just kind of [43:18] emerged to him [43:22] so let me go back over here [43:29] so here's our little toy world again um [43:37] now i want this to feel like a very [43:38] clock-like toy world so you'll notice it [43:41] actually is very exaggerated [43:42] um but in a kind of a stylistic way i'm [43:45] gonna go back to our little city here [43:46] see how they're doing [43:48] i think we have a crater where that [43:49] building used to be [43:51] in the vehicle editor there's one more [43:53] slot that over time we could eventually [43:55] unlock as we're playing through the game [43:56] um you know we started with ground [43:58] vehicles water air underwater [44:00] um and at this point uh i can now buy [44:02] what we call the ufo [44:04] and i'm just gonna buy a stock one you [44:05] know i could have brought this in the [44:06] vehicle editor as well [44:09] now this basically becomes your swiss [44:11] army knife at this point you know this [44:12] becomes the representation of all your [44:15] kind of tools and abilities in the game [44:17] this guy is very fast very mobile you [44:19] can go anywhere on the planet [44:20] in no time at all uh now there was [44:23] another landmark that really um [44:25] kind of inspired me at this level uh [44:28] just a little short side story here [44:29] um there was this show i used to watch [44:31] as a kid all the time it was called the [44:32] mutual of omaha's wildlife kingdom how [44:34] many of you saw that show oh yeah [44:38] so hey marlon perkins and this other [44:40] dude i forgot his name [44:41] but uh they would go around every [44:44] episode was them saving some wild [44:45] endangered animal [44:46] you know or some or studying it or [44:48] something but what i loved about the [44:49] show was that every week almost the way [44:51] that they [44:51] chose to solve the problem their [44:53] environmental problem almost always [44:55] involved a high-powered tranquilizer gun [44:56] in a helicopter [44:59] and so i watched the whole show just for [45:01] the scene at the end where the guys in [45:02] the helicopter hanging out trying to get [45:03] you know going through the herd you know [45:05] having the guy um so that was something [45:07] that i really wanted [45:08] uh with my ufo here um [45:12] and so the ufo really one of the things [45:14] it is it's a content creation [45:15] uh tool so um i can go around looking [45:18] for something [45:19] um let's see oh these are the guys that [45:21] were beating up us on us [45:22] way back now it's time for us to get our [45:25] revenge [45:29] okay if i can get over to him this is a [45:31] little tricky [45:33] got him yes so here this is basically a [45:36] little abduction beam [45:37] um and oh that guy's too fast for me [45:41] oh i don't want to tank no way i get [45:43] that guy oh [45:44] man they're hard to get well i guess one [45:46] will do [45:47] um so i can use this to go around the [45:50] world [45:51] sucking up content um and as you might [45:54] have guessed i can also [45:55] use my ufo to kind of go further out and [45:58] further out and further out [46:00] and further out and further out [46:05] [Applause] [46:13] so here's our little local solar system [46:15] um [46:16] a few comets there's our sun uh we have [46:19] hot planets near the sun [46:20] uh this is our home planet that we [46:22] actually evolved on [46:24] um and then we out here we have some [46:25] cold planets um we can click on any [46:27] planet kind of go over to it [46:28] and we can actually take our ufos there [46:30] to investigate [46:32] um these planets really represent like [46:35] little sandboxes for the player [46:37] so as we pull up to this moon you notice [46:40] geologically it's very different [46:42] than where we came from than our home [46:44] planet um [46:46] it's much further away from the uh sun [46:51] now one of the things i want to do with [46:52] this thing is i actually want to kind of [46:54] bootstrap it [46:55] um up to life you know and you know [46:57] basically build a whole flourishing you [46:58] know [46:59] civilization on this and i might do a [47:00] very different style too so you can view [47:02] these as like different palettes [47:03] that i have to work with um so one of [47:05] the things i want to do is i'm going to [47:06] try to get life [47:07] on this planet to live and kind of [47:09] bootstrap it up i'm going to come down [47:10] to the surface here [47:12] really quick um now [47:16] i'm going to try dropping that guy i [47:17] just picked up [47:29] um can anybody tell me why that happened [47:35] um if we can bring up the volume a [47:38] little bit on this machine [47:39] um yeah that's good um [47:43] so we had a little problem there uh and [47:45] we have no atmosphere [47:47] okay so what we really want to do is we [47:49] want to um [47:50] make this world livable now [47:54] at this level another one of my kind of [47:55] landmarks was an old thing called kid [47:57] picks how many of you played kid picks [47:59] um [48:03] the uh i wanted this to feel like kid [48:05] picks basically what i loved about kid [48:06] picks is you had all these cool weird [48:07] wacky creative tools [48:09] and they were always out of your control [48:10] i mean you always did these unexpected [48:11] oops i didn't mean to do that type of [48:12] thing [48:13] and i wanted the terraforming to kind of [48:14] feel like that as well um so i'll pick [48:16] one of my terraforming tools [48:17] and these are things that you earn for [48:19] the ufo you're actually increasing your [48:20] ufo's ability [48:21] over time and there are lots and lots of [48:23] tools and different ladders so um [48:25] we'll fire a little volcano tool [48:30] and so this is actually sculpting the [48:31] surface of the planet directly [48:35] and pumping atmosphere in there [48:49] now over time these will actually start [48:50] developing a thicker and thicker [48:51] atmosphere on the planet [48:55] [Applause] [48:57] but in the meantime i can actually start [48:59] building colonies here even while i'm [49:01] waiting for that [49:02] so took a spot zoom in [49:06] um when i build a colony now at this [49:08] point i get kind of the style that i had [49:10] on my homeworld [49:16] it has a bubble over it um this by the [49:18] way is also what i would get if i were [49:19] to build at this point [49:20] uh city underwater um so at this point [49:23] this level of technology i can start [49:24] colonizing [49:25] the oceans on my home planet and they [49:27] would have bubbles as well [49:29] um now this kind of takes a while now [49:31] once i terraform the planet i won't know [49:32] the need to build bubble cities and i [49:34] can actually build very elaborate kind [49:35] of colonies [49:36] um they're earning me money to help me [49:38] upgrade my ufo that's kind of the real [49:39] loop going on here [49:40] um we have a little cheat here you know [49:42] this is at the very high levels of the [49:43] game [49:44] um you'll eventually be able to afford [49:46] the we'll call the genesis device [49:50] so this will very rapidly if you notice [49:53] the ground texture [50:04] you see things are starting to green up [50:05] a bit we're getting a little bit of [50:07] atmosphere um [50:14] as the storm clouds pull back a bit um [50:19] so we're actually trying to bootstrap up [50:21] the atmosphere and then there's the [50:22] biosphere score so you [50:23] once you get a little bit of an [50:24] atmosphere you start then bringing in [50:25] plants probably that you've you know [50:26] harvested from your home planet [50:28] um and then you can actually start [50:30] living there so it's like kind of a [50:31] three-way colonization game terraforming [50:32] game [50:33] but really we want players to be [50:34] extremely creative kind of with what [50:36] they build here [50:37] um so let's pull away from this guy [50:41] so now it's kind of green that's a happy [50:42] little place we have our guys there [50:44] um so everything's good um as i [50:47] mentioned the ufo has a lot of different [50:49] uh [50:50] levels that you know you can advance on [50:52] the different tools that you add [50:53] one of the more expensive things that [50:54] you really have to kind of work in here [50:55] for a while [50:56] and build some usable colonies and a lot [50:58] of money but eventually you earn [51:00] the interstellar drive and so we pull [51:02] out again [51:16] now this is our home planet here um we [51:18] call it we call this the stellar zoo in [51:20] fact we want to have all [51:22] the real things that you would see in [51:23] space one thing that's always really [51:25] annoyed me about the way astronomy is [51:27] presented [51:27] is it's presented as two-dimensional if [51:29] you look at star maps like that i mean [51:30] that's the way astronomy is taught [51:31] but it's yet it's this beautifully [51:33] three-dimensional thing and you just are [51:34] not given the sense of what that's like [51:37] um we have these beautiful images from [51:39] hubble and we're never given any context [51:41] for how big they are [51:42] where they live how frequent we see them [51:44] in space and i really wanted to [51:46] at this level give a player a real sense [51:47] of what it feels like to be out in the [51:49] galaxy [51:50] so we have things like black holes [51:54] back here we have planetary nebula [51:57] and this is basically an old supernova [52:00] shell these very bright stars are [52:04] like blue giants that in fact are going [52:05] to go supernova fairly soon so i have to [52:06] kind of stay away from those so there's [52:08] actually a terrain here [52:09] to this this is the mysterio nebula back [52:11] here where new stars are formed [52:13] um but also at this level [52:16] we uh have a seti tool i can roll the [52:19] mouse over different planets and listen [52:20] to them [52:24] if i i'm listening for radio chat or i'm [52:26] actually trying to discover other [52:27] intelligence [52:28] oh there we that go [52:32] there we go yeah it's in an alien [52:35] language but i now know that they're [52:36] aliens living there [52:37] so um we can go visit that star now [52:41] so this is a different system that we're [52:42] in um looks like they had a little [52:44] accident [52:44] or something here um [52:47] this is another race now this is the [52:49] point at which i am overtly browsing [52:50] other players worlds [52:52] okay so this entire world was created by [52:55] another player and now the computer [52:57] you know it's downloaded to my machine i [52:58] want to emphasize this is asynchronous [53:00] um it's not a you know simultaneous [53:02] multiplayer game i downloaded it from [53:03] the other player [53:04] and the computer takes over control [53:06] based upon the personality sliders that [53:07] the other players kind of set for it [53:09] when they evolved it [53:10] um so let's go visit these guys [53:16] this is kind of a little hot planet [53:18] because it's so close to the sun [53:22] you can see it has kind of a very [53:23] different geology [53:26] very arid very large sun there [53:34] looks like we have a big city over here [53:36] we'll kind of approach slowly [53:37] now when we're doing uh first contact [53:41] with the other species um you have to [53:43] kind of play this little close [53:44] encounters musical game with them to try [53:45] and establish a language [53:47] um and they may or may not respond well [53:49] depending on how you play it [53:50] so i'll try that now [53:57] [Applause] [54:03] okay but luckily we have defensive tools [54:05] as well [54:17] so this is space invaders [54:20] uh my weapons are seeming a little [54:23] ineffectual against them uh but as i [54:26] mentioned you know all of our [54:27] you know tools on the ufo come in many [54:29] different flavors um [54:36] so [54:41] [Applause] [54:50] uh [54:57] so there you go our friends and there [54:59] goes my reputation as a non-violent game [55:01] developer [55:05] but we made a new ring it was a nice [55:06] crafts project [55:09] you know um so let's pull out to here [55:12] we're gonna leave this place there's [55:14] nothing more to see here move away now [55:16] um okay so [55:21] at this level the space level we [55:22] actually want to build a set of meta [55:24] games [55:24] around you know basically my favorite [55:26] genres of science fiction each one of [55:27] these is going to be a meta game at this [55:29] level [55:30] um you know starting with something like [55:31] war the world um [55:33] kind of an adult supervision peacemaker [55:36] day of the year stood still [55:37] type scenarios um uplift where i go to [55:40] worlds that have just primitive [55:41] creatures [55:42] i can drop down a monolith bring you [55:44] know and basically come back later and [55:45] see how they've evolved intelligence [55:46] to see those worlds um first contact [55:50] uh abduction and cross-breeding of [55:52] course you know you will have free [55:53] access to the uh [55:54] to all you know most of the editors that [55:55] point you know so at this point i can go [55:57] back into the [55:57] creature editor for free and make weird [55:59] alien you know hybrids [56:01] um and diplomacy uh every race is gonna [56:04] have its own kind of personality [56:06] um we're actually using the star trek [56:07] races as kind of landmarks [56:09] making sure that we get you know with [56:10] the smallest number of dimensions we can [56:12] clearly distinguish between all these [56:13] races [56:14] so that when i go up and interact you [56:15] know and i have like the klingons versus [56:17] the borg [56:17] or my versions of these things that they [56:19] should behave appropriately with [56:20] each other you know these when i want to [56:22] fight or trade or [56:24] just be good friends and share you know [56:25] cool ideas whatever um [56:27] we want the space game to really feel [56:28] very very open-ended and this is where i [56:30] think most of the narrative is going to [56:31] come into the game [56:32] um i want the players to feel like [56:34] they're creating these worlds if you [56:34] look at some of the old comic book [56:36] covers from the 30s and 40s [56:38] um that's what i want the players to [56:40] feel like they can make these worlds [56:41] worlds of this variety [56:42] if you look at these covers they mostly [56:44] you know they have vehicles buildings [56:46] aliens and planets i mean so all the [56:48] different things we have editors for are [56:49] represented on these covers [56:51] um this is some of the concept art we [56:53] did for trying to get a sense of how [56:54] broad we think players can build worlds [56:57] aesthetically [57:04] now the pollination at this level is [57:06] that we're actually doing overt browsing [57:08] you know i'm actually going to another [57:09] player's entire planet [57:11] when i'm on the planet i can actually [57:12] you know abduct their creatures steal [57:14] vehicles and they enter my library i can [57:16] use their creatures to go back [57:17] and populate like an entire zoo planet [57:19] if i choose to [57:21] so really we have these different levels [57:23] of gameplay [57:25] and the overall structure of the game [57:27] now is that of a t [57:29] um that we're going up through evolution [57:31] through all these different levels [57:32] and then we have these very broad set of [57:35] metagames [57:36] above it at the space level and this is [57:38] actually a little bit opposite than most [57:39] games because what we're doing here is [57:41] we're putting the more goal oriented [57:42] gaming first [57:43] as the tutorial for this the eventual [57:46] sandbox [57:47] um which will then surf down through all [57:49] these levels most games you think about [57:50] the sandbox is kind of the tutorial [57:52] that's when you learn to play the game [57:53] and then you go play the real [57:54] goal-oriented game [57:55] we're using the goal-oriented game to [57:57] basically train you to use all the [57:58] editors [57:59] and to teach you the simulation dynamics [58:00] at every level so by the time you get to [58:02] the space game [58:03] you can kind of fractally surf you know [58:05] both horizontally you know from planet [58:07] to planet [58:08] and vertically you know from like city [58:10] to sieve to tribal [58:15] so eventually um you know with the [58:17] players creating [58:18] all these different worlds here you know [58:20] a little bit more [58:27] now the region we were in was like this [58:28] little square here [58:30] um you know i think we can actually [58:33] create [58:34] a whole galaxy here of you know [58:37] tens hundreds of thousands of worlds [58:40] that the player can visit and interact [58:42] with [58:42] and kind of make up their own game along [58:44] the way um [58:45] this was you know you might kind of [58:47] notice this was an idea i've had a long [58:49] time ago was powers of ten i've always [58:50] been a big fan of the eames [58:52] um but i never in my mind could figure [58:54] out a way to kind of roll this into a [58:55] game and actually make it all add up um [58:57] more recently i started working on a [58:59] seti game um [59:01] because i was very fascinated with the [59:02] seti and astrobiology programs one thing [59:04] i noticed at some point though was that [59:05] the [59:05] interesting terms of drake's equation [59:07] and this is the equation about [59:09] you know a galaxy has this many stars [59:10] and this percentage of them have planets [59:12] and this percentage of them have life [59:14] um the terms in that equation map very [59:16] neatly to the different powers of scale [59:19] and so at some point i rolled the two up [59:20] the city game and the powers of tin [59:22] thing [59:23] this is actually one of the first design [59:25] dock which uh [59:28] wired about a year ago asked me to do a [59:30] page for them they said we're doing an [59:31] issue [59:32] on life and will you lay out a page for [59:34] us and i said sure can i do whatever i [59:35] want [59:36] and i said yeah and i just sent them [59:37] this i didn't tell them what it was [59:39] and they published it last year so the [59:40] design docs have been out and wired for [59:42] a year now [59:47] um so it's a very simple player story [59:48] here um and we needed a way to kind of [59:51] you know [59:52] compress this whole idea so that people [59:53] can comprehend it is that the player is [59:55] going from bacteria [59:56] to a galactic god um [60:00] and that we're kind of basically trying [60:02] to you know replay you know the complete [60:04] history and future of life [60:06] um which it's hard to come with a really [60:08] kind of [60:09] catchy marketing line for that if [60:10] anybody comes up with one let me know [60:12] please [60:13] um now the the hub that all of this [60:16] revolves around [60:17] um and i haven't had much time today to [60:18] go into the details but the hub this [60:20] really revolves around his compression [60:21] as i mentioned um the compression gives [60:23] us tremendous leverage for the player [60:25] um to make them creative which is a huge [60:27] win for the player [60:29] um it also the same leverage gives us [60:31] generative systems [60:32] um which we can use to populate huge [60:34] content libraries [60:36] um also it makes the content very [60:39] portable [60:39] that we can use to build an unbounded [60:41] world around the player which is again [60:43] is a huge win [60:44] we did over 200 prototypes of this in [60:46] fact i've shown many of the prototypes [60:47] for this over the years [60:48] i just didn't tell you what they were [60:50] but um prototyping has become kind of a [60:52] religion for us around this [60:54] and one last thing i wanted to say here [60:56] uh as we're working on this game [60:59] you know a lot of the games i've worked [61:00] on i've always had obstacles usually [61:02] trying to you know to convince other [61:03] people it's a good idea or be sellable [61:04] or whatever [61:05] um when i look back on this idea you [61:08] know the biggest obstacle i truly had [61:10] was making myself believe that we could [61:13] build the game it was my own imagination [61:15] was the biggest bottleneck um once i [61:18] truly believed that this game was [61:19] buildable [61:20] it proved to be actually quite easy to [61:22] con i mean to persuade the rest of the [61:24] uh [61:25] of my staff and the executives and [61:27] everything that we could do it [61:28] um so i'd encourage all of you if you've [61:30] got like some totally weird idea [61:32] that is just so far outside the box that [61:34] you think you know there's no way that [61:35] would work [61:36] go back occasionally and revisit those [61:38] ideas because you know you just never [61:39] know [61:40] you know where they might lead to and uh [61:43] last thing is i can't talk about this [61:45] again until e3 i'm sorry i got special [61:46] permission to talk about this today [61:48] here um but i can't say a word of it [61:50] outside this hall [61:51] after this until e3 but that's it thank [61:54] you very much [62:18] you can keep clapping [62:21] [Applause]