Fate of the World, however, comes closer than anything else so far to an honest-to-goodness world sim. It has quite a bit of what one would want -- global politics, environmental crises, resource limits, technological breakthroughs, biodiversity dilemmas, and more. I'm really excited to try it out -- it's currently in beta, but purchasing it now gives access to the beta versions as well as the final version upon release.
Fate of the World is a dramatic global strategy game that puts all our futures in your hands. The game features a dramatic set of scenarios based on the latest science covering the next 200 years. You must manage a balancing act of protecting the Earth's resources and climate versus the needs of an ever-growing world population, who are demanding ever more food, power, and living space. Will you help the whole planet or will you be an agent of destruction?
My big question about Fate of the World is whether it is a (for lack of a better term) first-order simulation, where events happen as direct results of the rules embedded in the code, or a second-order sim, where events happen because of the interaction of basic environmental conditions and player actions. First-order sims are straightforward and fairly robust -- the player can't do anything the game doesn't explicitly allow. Second-order sims are much more complex, and as a result can be much more prone to "breaking" by producing nonsensical emergent results -- but also much more open to innovative solutions.
A basic example of a first-order game would be a classic text adventure, where there's usually only one way to achieve a result, even if your character holds multiple objects that would also, in reality, also work. For example, if you had to press a button to open your cell by throwing your shoe to hit it, only your shoe would work -- even if you had a brick, a book, or a frying pan (all typical text adventure supplies).
In a second-order version of the same situation, the game would "know" that shoes, bricks, books, and frying pans were all smaller objects, and that one thing a character could do with a smaller object was throw it. It might also allow you to throw the pillow from the cot in the cell (also a smaller object), but may have basic world rules that say that pillows are "soft" and don't hit hard when thrown. Conversely, it may also know that a brick is "very hard" and could damage what it hits.
Much more complex to program, much more prone to weird outcomes, but much more open to novel strategies (e.g., throwing the pillow over the button, then throwing the brick onto the pillow).
We're still a ways away from being able to build fully second-order global simulations. It's not just going to require a lot more processing power and much more data to pull from, it's going to require much better models of underlying systems, models that can interact without leading to weird emergent results.
The worry I have about this surge of interest in games is that people who aren't familiar with the reality of games and simulations, only the Hollywood-esque version where every computer has a Do-What-I-Mean interface and every simulation perfectly captures reality, are going to expect much more than they get. Disappointment with the mundane limits of real games may mean that interest in games crashes just as quickly as it arose. I hope not, but it's incumbent upon us who do understand what games and sims can and cannot do to make sure we explain this clearly to our new audiences.