SD-cast
SD-cast
“What is your SDory, Karim Chichakly?”
Karim Chichakly is co-president of isee systems and an adjunct professor in the Social Science and Policy Studies department at Worcester Polytechnic Institute. He teaches system dynamics courses. He earned his PhD in Computer Science and Complex Systems from the University of Vermont.
Transcript of the interview: http://bit.ly/SD-cast-Ep3-Transcript
I would also like to thank isee systems for developing products, such as, Stella Architect and the isee exchange and providing services, such as, teaching to aid those who use system dynamics from beginner to expert levels. Below is a poem I wrote about Karim.
Dr. Karim “The Dream” Chichakly
Karim, “The Dream”
Is someone you’d want on your team
To model, manage and code
Guide you along the road
Of learning system dynamics
He is calm when others are frantic
He is a tough but fair grader
Will push you to do greater
Like his team at isee systems
Who are full of SD wiSDoms
See below for the links to Karim’s and Bob’s COVID presentations and models:
Karim Chichakly’s COVID-19 Model by Age
Karim Chichakly’s Vertical Isolation in Brazil Simulator
Karim Chichakly’s COVID-19 Economy Model
Karim Chichakly’s Modeling COVID-19 Policies (Video explanation of Vertical Isolation and Economy Models)
Bob Eberlein presents "Understanding the Unknown: Trying to make COVID 19 discussion as transparent as possible”
Bob Eberlein’s COVID-19 Simulator
Please subscribe to SD/ST-cast to hear more SD/STories.
Email me, ctang@wpi.edu, if you would like to be on SD/ST-cast or recommend someone who would be.
See below for the WPI SD Social Media accounts:
https://twitter.com/WPISDclub
https://www.linkedin.com/groups/1916314/
Music:
“Limelight” by Podington Bear is licensed under the Attribution-NonCommercial 3.0 International License.
I cut the music track from 0:465-1:20 and kept the next 6.85 seconds then moved the rest of the song to the end of the podcast and trimmed the end (~ 21 seconds off) to fit the podcast ending
https://freemusicarchive.org/music/Podington_Bear/Haplessly_Happy/Limelight
Hello, SD-cast listeners. My name is Christine Tang. I am an Interdisciplinary PhD Student in System Dynamics (SD) at Worcester Polytechnic Institute (WPI) and am the WPI SD Social Media Manager. In this podcast, I will interview someone in the System Dynamics/Systems Thinking community. This series is called “What is your SDory?”
This is Episode 3. Titled: “What is your SDory, Karim Chichakly?”
Biography
Karim Chichakly is co-president of isee systems and an adjunct professor in the Social Science and Policy Studies department at Worcester Polytechnic Institute. He teaches system dynamics courses. He earned his PhD in Computer Science and Complex Systems from the University of Vermont.
Thank you, Karim, for joining us today.
Karim: Thank you, Christine, for having me here today.
Interview
Christine: We would like to know when and how did you ‘discover’ System Dynamics (SD) or Systems Thinking (ST)?
Karim: It was back in college that I discovered it. I was looking for something in the library and just stumbled upon Limits to Growth. It was near where--what I was looking for and the name, the title drew my attention. I started looking through it. It resonated with a lot of ideas I'd been throwing around and talking about with people. Then I later heard from a friend that [in] system dynamics they numerically solve differential equations which I was very interested in at that time and also that the Meadows were at Dartmouth so I decided to learn more.
Christine: And what was the first model you encountered? You already said Limits to Growth. Did you work on that or...
Karim:Yeah, so the first model I saw was the World-03 model in Limits to Growth of course but I didn't spend a lot of time examining the model as much as I spent reading the text and the ideas in the book and the results of the simulations at that point.
Christine: Did you take a course? What was the first model you built in that?
Karim: Yeah, so I took, because of this experience, I took a course with Dennis Meadows in system dynamics modeling and it's strange, in the middle of this pandemic, to remember that the first model I built was an SIR model of [Christine and Karim laugh] that season's influenza and that was written in DYNAMO. I did learn an important lesson from that model. I had some problems because Dennis had given us data for the season and we were supposed to build a model and the model was supposed to match the data. It did not match the data. I spent a lot of time trying to get it to match the data--changing the model. I spent a lot of time researching before I even built the model. I had parameters that were set based on expert opinion etc etc and I couldn't get it to match and Dennis told me to change the parameters [laughs]. You have to remember I was a scientist at this point. He actually used the word fudge which--it was not a great choice but he was kind of surrounded by students at that moment. But I had a strong scientific background and I was at the engineering school at Thayer--school of engineering at Dartmouth and we called the people who were doing system dynamics the pseudo-engineers and when he suggested that I change a parameter that had been discovered through scientific study that seemed very wrong to me. [Karim and Christine laugh]
It wasn't until later. Two years later, Barry Richmond started. I was his TA and he told me, this is very important. (I’m sure Dennis would have told me if he had more time--if he wasn't surrounded by students at the moment). He [Barry Richmond] told me that I have to remember that our models are not reality and while the parameter in real life might be 0.1. In our model it may be 0.2 and that we shouldn't put too much stock in the actual value of the parameter or model. We should start with what we think it is but it's okay to change it because we haven't modeled all the complexity of reality in our model. It's a simplification. I think that's an important lesson for everyone to remember.
Christine: Thank you and what are you currently working on?
Karim: When the pandemic started, like everyone else in SD, I built an SIR model to look at the different impacts of the policies that were being thrown around. As I said, many people were building models. Bob Eberlein built one that disaggregated by severity. He was most interested in the burden on hospitals and the whole question about flattening the curve. I disaggregated mine by age because I was more interested in the effects of different policies given different contact behaviors and virus effects in the different age groups. So, for example, late teens and young adults were less likely to follow the quarantine rules and if the schools were open the children were all mixing there and you could also figure out severity if you needed to because there was data that had severity by age as well. We released this finally when Bolsonaro started pushing vertical isolation which is the idea that we let everyone else go out but just the older people who are most at risk stay at home, so this is a very age specific policy. And at the same time there were protests in the United States where people were also pushing for vertical isolation and it'd all be fine but the model--the model clearly shows this is disastrous because of how mixing works and how people live together in houses, etc. The other thing I was looking at the time was the impact of the shutdowns on the economy. I was looking at business closures and what was happening to unemployment. I built a model to look at both of those and I looked at the policy--The Cares Act--to see what would happen and the Paycheck Protection Program in there had a very large impact. It was a very good thing. According to the model anyway. People ask me whether I plan to tie those two models together and I originally imagined I might but I haven't because I think it's a fool's choice to try to choose between the virus and the economy.
First of all when people are making that choice they're not looking at the huge physical, psychological and economic cost to the virus running out of control but also the economy is going to be stifled if people are getting sick and if they're afraid of getting sick or dying and they're not going out so it's really a fool's choice. You need to get the virus under control first. That's the most important thing. So I haven't worked on those in a few months but…
Currently, I'm working on a model--I'm trying to build a model of structural racism. It's pretty complicated. I'm having a hard time paring it down. It's getting bigger as I look at it but I want to try to keep it simple and still be realistic. The model's based on one of the group model building projects I facilitated in Brazil. We'd come up with a map at that point and I'm trying to turn that into a model.
Christine Thank you and we will add links to the presentations you gave and the isee systems exchange links to your models. [Karim: Thank you} So what do you suggest novice modelers should do to help in this pandemic?
Karim: I think the biggest thing that anyone who's familiar with modeling can do is to try to help people become more informed. There's so much misinformation out there. There's so...there's still people out there who say “There is no pandemic. This is all hoax.” [Karim and Christine laugh]
There are plenty of people who think you know using masks and social distancing don't do anything and there's no point in doing that. Anything that people can do to try to explain the physics of how this is happening, the actual, you know, the fact that contacts lead to infections and how that happens and how this grows exponentially so it gets out of control very quickly. Anything that you can do to help people understand the consequences of not taking this seriously and also the positive effects of doing the things, you know, on the other side--the positive effects of doing the things that have been suggested
Christine. Okay. Thank you. What other wiSDom do you have for students and those new to system dynamics/systems thinking?
Karim: I think it's important that you learn a field outside system dynamics and then apply system dynamics to that.
Also a lot of people get caught up in the idea that system dynamics is the one thing to solve everything. Don't get caught up in that. There are other methods out there and it's important to choose the appropriate tool for the job whether that is system dynamics or it's agent based modeling or discrete event simulation or whatever you want to try. Just remember that there are multiple tools out there for the complex problems that we're trying to solve.
I think it's important when you don't know a subject well that you work with subject matter experts. You should use data whenever you can find it and whenever it's possible and you should also follow the field's best practices, for example, using known components or molecules.
It's important very very important this is probably the most important thing--never fall in love with your model. [Karim and CT laugh] It's all too easy after working on it for a while to think your model is really great and wonderful and it is, I'm sure, in some ways--always. The problem is it's not reality. Just remember it's not reality and may be how you perceive it or what your research has shown but other people may have perspectives that will show that it's different. You also may get in a situation where it's showing anomalous behavior and it's because there's stuff missing from your model that you're not aware of so just don't just don't fall in love with your model. Be aware that you may have to change it. You may have to throw it out. Barry Richmond. It's it's really kind of funny. Barry Richmond. He used to write a model. He’d build a short model. He’d go “this is great” and he'd share it with everyone and he'd explain it and he'd run simulations. The next day it was all ripped up [Karim and CT laugh] and he'd built a new model. He's like “that one was wrong” and he'd do the same thing again and the very next day he'd rip that one up, literally, because he'd have printouts and he'd say, “that was wrong too.” He had no problem throwing away a model and starting all over from scratch. I think that's a really good model for how (model?) for how we should behave when we're working on our own models.
The last thing I want to say is, as Steve Peterson always says “Never stray too far from a running model.” So as you're building your model make sure you can run it in pieces even if everything isn't there yet to make sure everything makes sense at that point.
Christine: Words of wisdom and humility from one of the co-presidents of isee systems...so do you have any funny or fun SD stories that you're willing to share?
Karim: Yeah, so there have been a lot of funny things that have happened over the years and there's been a lot of fun as well but I'm going to stick to just three short things.
The first one's kind of an inside joke, so you may not get it; but, Barry Richmond, he always used to make fun of some of the terminology in the [19]90s...so people would talk in very, you know, very complicated words when they could. So a very common thing that people would say [is] “the overarching goal is to solve these underlying issues.” Right? [Karim is laughing while telling the story]] Something like that and so he would, you know, when he'd say overarching he'd he he'd with his hand he'd draw an upside down ‘u’ and then when he'd do underlying he would draw an up--he'd draw ‘u’ with his hand. Right? And whenever you talk about something and it wasn't really clear or someone didn't understand. He'd always jump in and go “that's because it's both overarching and underlying.” It was a running gag he'd do this all the time or even when you did explain something and uh you'd say someone would ask well how could that be so “because it's overarching and it's underlying” so that was kind of an inside running gag that went on for a long time
Another thing that I still laugh about whenever I think about it: Khalid Saeed in his SD 551 classes a long time ago in the lectures...he tries to stay very professional [in] the ones that were that were filmed in the studio and very uh straight and you know he's talking like this is a very serious subject that he's talking about. All the time very very serious, you have [Karim and CT laugh] if you have any doubts about the seriousness, just look at his face. He's very serious about this um but there's one point in the course where he's trying to tell a very funny story. Okay? But he's trying to maintain that seriousness and the whole time you can see the side of his mouth twitching up as he's trying not to break out laughing while he's telling this very very this very funny story in a very very serious way and whenever I think about it, I can't help but laugh.
The last one has to do with uh one of my trips to Brazil when I was facilitating a group. I was actually teaching them systems thinking. I had a class of 90 students and I used some of the things from The Systems Thinking Playbook. I don't know if you're familiar with that but that has a bunch of games that help people understand the concepts better and I had also invented some of my own. I designed some of my own games to augment what was there because there weren't games for everything I wanted to do and one of the ones I designed was for Accidental Adversaries.
This is when two people are trying to work together in partnership but then do things that hurt the other partner inadvertently because they're looking out for their own interests. They weren't thinking about their partnership. In any case, I'd set up groups around the room to work on this game and go through this game and all of a sudden a woman stands up and very loudly says “You betrayed me! You betrayed me! I can't believe it! You betrayed me!” [Karim and CT laugh] It was hilariously funny. I came over there and she goes “He said--he promised me that he would do this and he didn't! He betrayed me!” She was outraged in the middle of this game that he would do something...well, I mean this was the whole point, of course, accidental adversaries. He was doing something to protect his interests without realizing it was going to hurt her interests but...the game was successful but it was a funny moment. Stopped the whole class.
Christine: Did you ever publish these games or are the ones you created online somewhere?
Karim: No, I haven't. I have them all in my files.
Christine: Do you ever plan to?
Karim: I had never thought about it. I suppose I could. I could, you know, somehow make them available. Some of them are just on paper games and some of them require a model to go with them which I developed.
Christine: Okay. Thank you so much for joining us today and sharing your experience.
Karim: Thank you for having me. It was a pleasure.
Ending
I would also like to thank isee systems for developing products, such as, Stella Architect and the isee exchange and providing services, such as, teaching to aid those who use system dynamics from beginner to expert levels. Below is a poem I wrote about Karim.
Dr. Karim “The Dream” Chichakly
Karim, “The Dream”
Is someone you’d want on your team
To model, manage and code
Guide you along the road
Of learning system dynamics
He is calm when others are frantic
He is a tough but fair grader
Will push you to do greater
Like his team at isee systems
Who are full of SD wiSDoms
See below for the links to Karim’s and Bob’s COVID presentations and models:
Karim Chichakly’s COVID-19 Model by Age
Karim Chichakly’s Vertical Isolation in Brazil Simulator
Karim Chichakly’s COVID-19 Economy Model
Karim Chichakly’s Modeling COVID-19 Policies (Video explanation of Vertical Isolation and Economy Models)
Bob Eberlein presents "Understanding the Unknown: Trying to make COVID 19 discussion as transparent as possible”
Bob Eberlein’s COVID-19 Simulator
Please subscribe to SD/ST-cast to hear more SD/STories.
Email me, ctang@wpi.edu, if you would like to be on SD/ST-cast or recommend someone who would be.
See below for the WPI SD Social Media accounts: