Skip to 0 minutes and 10 secondsThe experiments of Thomas Schelling show that you can show emergent processes using a very simple model. And this is exactly what you can do with an Agent Based Model. Demonstrate the principle. It's not about recreating the reality. Now, you want to demonstrate principles of emergence that's applies to society. And I followed the idea of Joshua Epstein who tells that, "If you grow the phenomena, you'll understand how it works." And understanding how it works, gets a deeper sense of knowledge than just using static cross sectional correlational-based analysis. No. You really want to understand how this process works.
Skip to 0 minutes and 56 secondsNow, one of the interesting things of Agent Based Modelling is that it allows for running many experiments in identical settings with many, many agents over long, long time periods. Things that are impossible to realise with real people in a laboratory setting. So Agent Based Modelling is perfect for studying human behaviour in complex settings such as ecological and economic systems that show developments over many, many years over many, many people. And instead of showing one outcome, they can show you landscapes of outcomes that are more likely to emerge and the outcomes that are less likely to emerge that might be a possible outcome.
Skip to 1 minute and 46 secondsWe can now experiment, how can we move towards a certain outcome that we like or how can we move away from a very nasty outcome. These are possibilities with Agent Based Models. And we can experiment with heterogeneity. Can we address very specific agents in our model? And can these agents, in turn, have an impact on other agents? So we have social influence running in a model. That's possible when we model people. And first models based on this, originate from what is called econophysics. And econophysics is a kind of science where a physicist thought about, hey, the models of atoms, why don't you think about social atoms and use these models to understand how new products spread in society.
Skip to 2 minutes and 50 secondsThey called percolation models. Basically, it's is the same as innovation diffusion models. And of course this process is very important, because change of society deals with innovation diffusion, how new behaviour, technologies spread through social systems. Think, for example, about electric cars. In the next step, I will show you a movie of how this has been worked out in econophysics.
Why make an agent based model?
This video explains why it is helpful to make an Agent Based Model and what the benefits are of creating a model for understanding human behaviour rather than recreating such an experiment in a laboratory setting.
Note that innovation diffusion is the rate at which new ideas and technologies spread through society.
© University of Groningen