I just got out of Will Wright's keynote here at SXSW in Austin; a packed ballroom, a gobsmacked crowd. We got a Spore demo - an on-the-fly, operated by Will Wright demo - during which we got to see Spore from tiny slug to entire galaxy. We saw a player-generated facsimile of the Star Trek Enterprise scoot about the universe, planting monoliths, dropping in on planets, creating ozone layers and greenhouse gases and planet meltdown. We skipped through star clusters, and watched stars form, visiting player-inhabited stars of neighbouring space.

Spore really is a universe. I hadn't quite grasped how BIG it is. It's as big as players will make it, and it's beautiful to boot. Will is, as I'm sure you know, an entertaining speaker; he skips about massive amounts of information and data with ease and clarity. You can see Spore's influences: The Sims, Black & White, Star Wars, Star Trek. Robert Heinlein maybe, Darwin of course.

Here: I took some notes, I'm reproducing them here so you can grab an idea of what it was like. Don't take them verbatim - Will's a lighting speaker - but it should give you a bit of a gist of what he said... Alice Taylor

All those pics you just saw from the Hubble I thought I'd inflict on you, and I broke my arm skiing, and I've had way too much coffee today so we're going really fast today.

I figured SXSW is this filmy, interactive thing, festival, so I thought I'd talk about story, then about a week ago I read the description of what I was actually supposed to be talking about, which is Spore ... so I've done a mashup of story and Spore for you here.

So a few thoughts on storytelling. First why .. I hate stories, stories that my computer tries to tell me. Story's been the model from movies, it's kinda our heritage. But first of all the nature of story... I look at the world as a simulation. Here's a world stage. Lots of things cascade into the next stage. And certain things cause changes in other things.

Story causes a chain and conveys it to a viewer... a story's all about the chain of events, very linear, unchanging, you've all seen the same version of Star Wars.

But games are very open ended. Also, movies are primarily visual. Games are primarily interactive. So when we take away the control from a player, we're taking away the most important thing from them. It's like going to the movies and showing a blank screen...

Different games have different levels of interactivity, in some sense... here's chess... [..] we're trying to generate the largest rulespace in a game. This is the opposite of science, where we try to find simple rules to describe all this data. There's this topological difference.

It's because of the POV. When you're telling a story in a movie, it's from a chosen POV, it's all controlled, but games, games look like this [screen of wiggles and randoms]. You go up here, you lose, so you go back to the beginning. Over here, you lose, try here. Back to the beginning. So movies are far more compelling than interactive drama, because interactive drama is quite flat.

But empathy is really important to me. Movies have these wonderful things called actors, which are like emotional avatars, and you kinda feel what they're feeling, it's very effective. Films have a rich emotional palette because they have actors. Games often appeal to the reptilian brain - fear, action - but they have a different emotional palette. There are things you feel in games - like pride, accomplishment, guilt even! - that you'll never feel in a movie. I felt so bad about beating my creature to death in Black & White.

Stories are about empathy, and games are about agency. I'm causing what's going on on screen. Can I do this? Can I accomplish this? These models are cognitive technology. They're the original educational technology. They involve abstracting the world. Both of these respond to being stuck in time, but we wanna move experiences outside time... so...

[slides]

A lot of stories start out with here are all these characters [Leia, Luke]... and the structure is pretty fuzzy. A much wider set of possibilities can unfold from the beginning of the movie; think about Star Wars. Causal links. Think of that scene from Indiana Jones and all these traps are going off, you're filling in all these possibilities yourself, what if he falls in the hole, what if , what if .. and that makes it interesting. Star Wars grips you in the realism of the backstory, the fact that the ships were dirty [tells you something]...

With linear stories you want to start amplifying. In Star Wars, it all comes down to these two possibilities - the rebels are crushed, or the Death Star blows up. One of the fundamental things I've found as an interactive storyteller is that in linear stories the director knows the future. He or she knows the minor details that are important to present to you. But we [interactive storytellers] don't know those things. Ours are chaotic systems. Very minor initial conditions can lead to wide-ranging end conditions.

In a linear drama you can show the causal chain; in interactive drama you can't so much. We're playing with it in movies in interesting ways: interesting sub-plots (Magnolia); one of the things that keeps your interest is your wondering how these back stories are going to come together. This is a common thing that people do with causal change. Timecode is another version of this. It's parallel multi-threaded storytelling.

One of my favorite types of movie is one that's going along, la la la, and then suddenly it takes a crazy turn and it's like holy shit! This is not the movie I was expecting to see...

Memento is really interesting causal change. As events unfolded, each point caused you to re-evaluate what happened before. You had to reconstruct what happened - it's like a puzzle game. One of my favourite gamelike movies is Groundhog Day. You have this sequence, and then... it's back to the beginning. And it happens again. And again. It was a really interesting example of the director knowing the future as well as the past. Every day you'd seen the differences... you'd cover an eternity of experience.

This is something we really should be doing with games, if a player has failed on the same miserable level three times in a row, shouldn't we let them skip that level?

I'm going to attempt to try to tell you about game stories and player stories, which I find far more interesting. You may be familiar with CAVE OF TIME books. Here's this Branching model. But this is very expensive. Too many branches. Here's a Gating story: Quake's kinda like this. You can do what you want, as long as you find the key to get to the next level. Hybrid is another model... Gated plus Branching... but a player will actually play a very small percentage of these trees. So this has been the downfall of these tightly topographical branching narrative stories. [They're expensive and inefficient.]

Here's an interesting version I want to present to you: Generated. You have story fragments, with triggers... you can put them together like Legos, and form a story over time. It sort of makes causal sense. It's a form of procedural storytelling. One of my favourite short stories of all times is Maneki Neko by Bruce Sterling.

There was this karma computer in this book. The more the protagonist obeys this karmic computer, the more other players would help him. I find this fascinating to think about with multiplayer, where other players could help you [...]

Media is malleable in this new generation. A computer used to be a fancy calculator, but nowadays it's really more of a communications device. So I think we are looking at technology as player-centered rather than broadcast-centered... the masses are creating their own cool stuff and they share it around with each other.

Players invariably come up with stories about what they did in games. They're never describing a cut scene. I categorise these as Unintentional, Subversive and Expressive.

Unintentional is when a player comes up and finds a bug, and they make a back story.
(e.g. spontaneous combustion in early versions of The Sims).

Subversive are where players are trying to push the envelope in different directions, exploits and cheats, etc. In Battlefield '42 you get coordinated cheat activities done as a group, filmed by players and uploaded.

Expressive are more like what we see in The Sims where players have an intentional message. Here's GTA: I spent my entire time creating a character, a semi homeless person hanging out with my homeboys and doing tricks on my bike. The Sims... people started playing it, and they'd be verbalizing the story as they played it. They were reducing it to a linear story - so we put up a web page for them to upload these stories, and we ended up with hundreds of thousands of them. Players became performers. The game became a storytelling tool. People were writing their own 'levels'. Machinima takes it even further.

[Player stories]

So kinda thinking about storytelling, looking at the computer and looking at entertainment, it's more about listening to a story rather than telling a story. It's about listening to the player stories; those are the ones they care about. We can map this. With neuro-linguistic programming you're taking a real sentence and decomposing it: there's something very similar in stories I think.

You can have the computer understand, "Oh I see, this is a boy meets girl story", etc. If we know what the goal states are, we can present dramatic obstacles, things to amplify the drama. The whole thing comes down to an epic struggle, perhaps. If we can parse the players intended story, we can change the lighting, the music... the events! If it's a horror story, we can add spooky music... we can add zombies. Maybe we drive events to clarify a story, and then actually you've created a movie. I think this [generative power] might happen by observing lots of parallel players and pulling the data out of that.

I'm thinking the Truman Show, where you would allow the player to run around with a certain amount of free will, and the computer is like the director, who controls the envelope around Truman but can't directly affect his movements. The Truman Show is basically a game. I wish games were more like the Truman Show. Can you bust out of the game? Groundhog Day is like a game for mapping out the possibility space around you...

There's this concept in games called the magic circle. When people play together, they sit and respect the rules. People outside the magic circle don't need to observe those rules. But if you enter it, you're agreeing to be bound by those conventions. Story's the same thing. It's a shared experience. Story circles started around campfires, and has evolved into more structure environments... we have these huge formal things in movie theatres, but it's shrinking back again with television. Shorter shows, in the living room, or even on video iPods. This is like fractal storytelling. I don't have to go to a movie theater or a living room to watch a story now, I can do it on the bus.

Interactive is riding this generational wave. It's a cultural overtake process. There's a lot of people who've spent a lot of time with linear, but this whole generation is coming up... who are more comfortable in interactive. Games are being thought of as a tool for self expression, like a hobby but also like a tool or ...

Players love making content in games. We've been riding that wave a lot. They love sharing and collecting content. Some people love just organising it. The power of that collective effort is amazing. You're seeing this on social network sites. But most of the content is not so good, and a smaller percentage is great, but as we give them better and better tools, we'll increase the quality of what they're doing.

Not only is it of value to them, but to others as well. So we can model the players ... we can understand what they like and what they're doing. How they move across the game play landscape. What they buy. We can look at social networks. Or even social interaction frequencies ... are they being mean, or nice, are they friendly? We can build elaborate models predicting this behaviour.

So with Spore... if we could give a player a tool to build a tinker toy, and the computer takes that and presents it back in hi-res oitput: this is creative amplification on the player's efforts. What if we could take those assets and collect them all on a server and categorise those, and then predict what they'd like and bring it back into their world? This is what Spore is like. I wanna take the player out of the protagonist of Luke Skywalker, and put them in the world of George Lucas. This is SETI, Drake's Equation... here: I'm going to show you a quick demo.

A lot of the work on Spore is about player tools. In the Sims it was about... you had to create the assets outside the game. There was lots of friction for players to create stuff. So the process in Spore of playing is the process of making assets. And we collect and redistribute these assets automatically as part of the game. We're able to build a kinda infinite sized world ...

Here's your little creature. He eats food pellets. He grows. We're going to transition between many orders of magnitude as we grow up. [Gets eaten] you finally move out of the water and onto land... so it starts like 2d Pac-Man, and it evolves... into a 3D 3rd person... here's my little guy... we're on land. We're a slimy slug thing. This world... at this point it's a really simple game of evolution: eat, survive, and reproduce. [Gets eaten]

Damn I wasn't supposed to die. Okay.

[laughter]

We want this to feel very tactile and toylike. This bit is really like play. Here's a mouth, here's limbs. I'm putting mouths on his arms. The game tries to interpret the player's intent. A lot of these parts have little morph handles... each part represents several hundred [states]. It has aspects of clay, Lego, Mister Potato Head.

So now I've created this 3D mesh in mere minutes, we've reduced 3 days of art gruntwork into a few milliseconds. Now we have to bring it to life, the computer will analyse where the legs are, how it dances, how it poses... we well as how it fights, eats. Stuff like that. The computer is dealing with mesh, texture, animation, all on the fly procedurally. And every different property is going to behave differently in the game.

We want the computer to be totally focused on what the player has made. It has to be front and centre... here, he's grown up a bit... [Skips ahead] so over time you end up managing whole tribes of your creatures, and civilizations... eventually they can build spacecraft. I want this game to bring up lots of interesting issues for the player. Where might life go? The future of life? The effect that life can have on the universe, it's philosophically staggering.

I studied Montessori's philosophy and methods. She basically wanted kids to explore the world themselves using toys and objects, learning the meaning of things... and I want to build a game where a player is going to come across the Copernican Principle, say, and you walk away thinking of the meaning of life. Or what the future might take.
[Wright chooses a Star Trek machine model]

One of my real aspirations of this is I wanna see interstellar wars between Care Bears and Klingons.

[laughter]

The cloud ray here increases the greenhouses gases... we can play out The Inconvenient Truth here. You can see oceans rising, cities destroyed... I can heat it up too and I can combat the rising levels by heating the planet enough to melt the oceans but that sort of kills off everything. My planet's melted!

[laughter]

So in some sense the entire planet is a toy.

One of the really nice things with a toy like this is you can give people long term dynamics over short term sense. It's so hard for us to think over the long term, longer than 100 years, but by using these toys we can help people to think and understand [..]]

See this planet. It has primitive life, as we play through evolution, we're trying to capture most of the dynamics of evolution. As we move into the future I want the game to focus on fictional landmarks we have. Lots of these levels are based on my favorite science fiction... here's my Monolith tool. I'm looking at my favorite sci-fi movies and figuring out what the landmarks are. So see... they've built a tribal society around this monolith, and let's see, maybe I can get them to worship me... I have to be careful. Hmm. I don't think he wants to worship me right now.

We have this idea of a "Sporeopedia". It's how we categorise everything you've seen so far in this world. This Sporeopedia builds automatically; we took stuff from x-files, star trek, war of the worlds. Eventually you can pull out into interstellar space around our star and we have lots of... look, Hubble objects. The player can go around like a tourist, visiting this little space zoo. It gives people context for things they're seeing. Here's an unexploded star, they're like the birds of the universe, these beautifully coloured things...

Every star system will have planets you can interact with, and many will be inhabited by other players' species. We're seeing thousands of stars... but this is a small fraction of the galaxy we're building here. So basically that is Spore.

[rapturous applause]

You can take any human technology and take it as a new extension of our body. Telescopes extend our eyes, cars our legs, telephones our voice. Computers do a lot of these things but the most important thing is that they extend our imagination.

This is a very powerful thing, an amplifier for imagination.

We use computers for entertainment, education, social spaces. How is this going to impact the world going forward? Every now and then the world goes through a huge paradigm shift... sometimes by social shifts, sometimes only once or twice in a lifetime. Some are grass roots, some are top down, and some take us by surprise.

We have a lot more heading our way. More political and social issues. Obviously environmental issues. Some people are issuing warnings.

But when we look at games specifically and entertainment in general, games often have this perception of mindless toys, but they can be much more than that. They can help us develop systematic thinking. They can help us build accurate models of the world around us; and hopefully these things will help us change the world just a little bit for the better.