SE vs. AE

While wandering through Mass Effect 2, I was struck with the vitality of the world. Circa 2004, the main attraction of a multiplayer environment relative to single player worlds was that single player worlds felt dead. Multiplayer, on the other hand, had vitality but also the annoyances of dealing with other people and their inevitable failure to be perfect friends, or perfect foils.That problem can be reduced by Social Engineering (SE): Designers use policy (sometimes enforced by code) to optimize an individual's experience when dealing with others. Judging from ME2, the problem of dead single-player worlds can be addressed successfully using a suite of tools involving digital storytelling, emotive animations, deep conversation scripts, and a strong responsiveness of the emotive/relational space of characters to the protagonist's actions. Altogether, let's call this bag of tricks "Artificial Emotion" or AE. It's not a new term, indeed Professor Turkle has paved the way here, as before.

As the market for fantasy evolves, these two approaches to improving happiness seem to be facing off.

Would humans be happier in an environment constructed just for them, when every other being behaves just so, even though none of those beings are actually people? Or would they be happier to live with other real people in an optimally designed social environment? Is it easier to improve AE or SE?

The issues extend far beyond the game industry. In this area as in others, the game industry is charting territory that business and governments will deal with soon enough. If developments in SE dominate those in AE, look for a future of massively-linked online communities whose policies produce much more happiness than offline communities. If AE wins, look for a future involving isolation pods. Most likely, we will have both. As for offline existence, SE advances might translate into better governance in the real world - better companies, better neighborhoods, better schools. AE advances seem less likely to help the offline world.

Life in the soon-to-be-launched Old Republic may combine the best SE and the best AE in one world.

Comments

[Mike Sellers]

1.

Interesting post Ted - both AE and SE are officially Hard Problems. Creating effective, resilient, non-brittle communities or effective, resilient, non-brittle representations of emotions in online worlds or NPCs are incredibly difficult to do. I still think they're worth doing, but I wouldn't look for them in mainstream games any time soon (I've been saying that for, oh, at least ten years, and it's still true.)

Personally I envision a combination of the two alternatives (points on a line?) that you mention. For example, I and others have for years been touting the possibilities of "pesonal NPCs" (probably going back to Mike Rozak's blog post on the topic from a few years ago).

If you can have NPCs that matter to you, and you to them, who are entangled with your life-journey in the game/space and have the ability to play with others... well that seems like a winning formula to me.

As for SWTOR, I have great hopes and significant concerns. Will all that necessarily static voice acting really stay fresh, or will it quickly become "tl;dr" as with quest text in other games?

[Richard]

2.

Ted>It's not a new term, indeed Professor Turkle has paved the way here, as before.

It wasn't a new term when she used it, either. I have papers on the subject from the early 1980s when I was doing my PhD.

Richard

[Lisa Galarneau]

3.

Isn't it highly possible that our universe itself is such, or some combination thereof (a handful of 'real' people, the rest NPCs)? Did you know that Philip K. Dick's middle name was 'Kindred'? Yet he was the guy most in doubt of the reality of our apparent SE.

I'd take AE as the whole kit and kaboodle might just be a big illusion anyway, and I don't need real people if AI is just is good. But that's gonna take a while:
http://www.wired.com/magazine/2010/12/ff_ai_essay_airevolution/

Yes, I have thought about this, kind of a lot. Would I take an android boyfriend that mimics Keanu Reeves, perfectly? Yes, yes, I would. I can turn off the body odor and snoring, you see, shut him down when I want control of the remote, and I wouldn't have to feed him, etc. Unless I want to.

[ecastronova]

4.

That brings up the whole Bostrom argument, that we're already in simulation.

[ecastronova]

5.

Richard> It wasn't a new term when she used it, either.

Nothing is new to you, Richard!!! It's like talking to Edison about light bulbs.

[Dean]

6.

@Lisa, I really love the reference to Philip K. Dick as you descend into a Cartesian conundrum. I have a problem with your Keanu Reeves android though: wouldn't the level of control you have over him/it expose the simulation for what it is. This may seem trivial, but surely the the simulation would need to seem convincing for you to engage with it in a genuinely emotional way.

[LisaG]

7.

Found this video on YouTube while reading up on Dr. Bostrom. (also something in there related to Ted's exodus idea).

Didn't realize Bostrom was so dystopian! Oh, is he the guy who says 'the universe is a crazy rampaging robot programmed to kill?' Maybe, but if I believed that, why would I get up each day?

[Lisa Galarneau]

8.

@Dean

The key, I think, is the equivalent of a romantic turing test that will tell if an android or bot is sentient enough to overcome my cynicism. If it appears substantially real and elicits romantic response in me and is sincere, what do I care if it actually is physically 'real' or not?

How far off are these kinds of choices? 10 years?

[Lisa Galarneau]

9.

Mmm, 5 years?
http://www.videosurf.comwww.videosurf.com/video/life-like-walking-female-robot-62356652

And Dean, realized I can answer your question better. While I mention control as a benefit, I wouldn't want complete control. Spontaneity, mystery and surprise fuel romance. They would all have to be complexly modeled and integrated into the AI.