John Sterman at the MIT Museum, reflecting on bathtub dynamics, public perceptions and political conflict over climate change:
Skip to content
Don't just do something, stand there! Reflections on the counterintuitive behavior of complex systems, seen through the eyes of System Dynamics, Systems Thinking and simulation.
It really was a fantastic talk. Can you guess who wrote the note during the QA period on putting it all together as a third political party? I hope John does a TED or TEDx talk soon. He’s such a great speaker.
I love the Roulette line too: “Basically what we are doing is playing Russian Roulette with the future for our children and our grandchildren. But unlike the classical game where you might use a revolver with six chambers, only one of which is loaded, here we are using a weapon where there are 19 of 20 chambers loaded. So I don’t know how you are feeling tonight, they won’t let me bring my AK to the museum, but spin the chambers (approaches audience member)… you feeling lucky?”
I’ve edited the video down a bit to cut out the slow parts and put together a toy model that does the basic bathtub dynamics (links below). I’d really like to convert the full C-LEARN model, but the special subscript functions are killing me. You don’t happen to have a lighter version you could share? Overall the C-LEARN UI leaves a lot to be desired, but instead of just criticizing I’d really like show you guys what would make it a lot clearer for people. The simple model should give you a taste of what I have in mind.
The edited talk:
http://www.thefoundationparty.org/climate_change_risks
My simple model: http://www.thefoundationparty.org/simple_model_of_climate_change
We’re beta testing a .dylib for Vensim on Mac OSX and iOS, soon Android, so it’ll be possible to build a cross-platform interface (e.g., Java) and leave all the UI stuff out of the model. That might be one way to do it. The UI issues with C-LEARN are mainly symptoms of the limitations of trying to do multi-modal stuff in Vensim, which is basically flat. Future Vensim will fix that. Travis Franck of CI has been working with Argonne on a Java translator for Vensim, which would be another way to go.
Yeah I just saw the iOS/Android announcement in the Vensim newsletter. I’ll be interested in seeing how that comes out. At my work we are actually separating the simulation from the UI as well via web services, but it’s still fairly code intensive whereas I can “prototype” things fairly easily with AnyLogic.
I did get the impression that a lot of the C-LEARN issues had to do with the options for choosing how many country sectors to use. I’ve basically settled on converting it from scratch so that I can pull things over into a more object oriented solution. I’ll be interested in seeing what you’re doing with Vensim to go beyond a flat IDE. If I do have questions on C-LEARN equations/functions, would you be the best person to ask?