This blog post tells a good example of my concept of learning. It is the most effective, as people look at the results using their eyes, comparing the results and reflect on the differences. Same input, different interventions, different outcomes. It must be due to the interventions! No cognitive conflict can be stronger than this. By explicitly pointing out this, again and again, people cannot escape but reflect.

If we have two earths (at least two human-hospitable planets), then we can run an experiment – one earth is like now, business as usual; and another earth changes to sustainable course. Wait and see them after Year 2050, we will learn that which one is better choice.

Unfortunately we don’t have another earth to learn. The best we can do is creating virtual earths, or modelling and simulation to test policy interventions. People who don’t like simulation outcome of a model can always complain about its assumptions, since model is never as perfect as the real world. But this is what the best human can do. Study the model as unbiased as possible, learn its usefulness and limitations, examine its assumptions and trace back to understand why the simulation outcome is like that (how the model results that simulation outcome). We can learn a lot even the model is not perfect.

Another way is to learn from history. Of course no history is identical. But similar to modelling and simulation, we can learn its usefulness (insights) and limitations (what may not apply). Understand why the history becomes like that is the most important gain of study history. Generally most of the insights are just repeating. If our civilization is collapsing again, it means we didn’t learn enough history.