On this episode of The Artists of Data Science, we get a chance to hear from Scott Page, a professor who studies complex systems and collective intelligence teams and political and economic institutions. He's known for his research on and modeling of diversity and complexity in the social sciences with a particular interest in the roles that diversity plays in complex systems. His book, “The Model Thinker”, stresses the application of ensembles of models to make sense of complex phenomena.
Scott shares with us his predictions into the future of machine learning, the importance of using a simple model, and how diversity impacts productivity. This episode is packed with amazing content that all data scientists and machine learning practitioners can apply in their lives. It was an absolute pleasure chatting with Scott!
WHAT YOU'LL LEARN
[12:41] Scariest applications of machine learning we might see
[24:56] What is a model, and why must they be simple?
[33:30] Many model thinking and it’s advantages
[47:07] How diversity impacts productivity
[49:46] How creativity impacts success, and how to be more creative
[6:31] “...you have to separate achievement from purpose.”
[35:45] “...if you really want to understand a complex phenomena, you've got to look at it with lots of lenses…”
[45:02] “...what you really want...is people who are acquiring different ways of thinking and understanding different tools, because then the whole is going to be so much more than the sum of the parts.”
[46:36] “Creativity is the union of sets. Getting at the truth is the intersection of sets.”
[00:01:15] Introduction for our guest
[00:02:45] What drew you to the field of modeling in general and specifically game theory and complexity?
[00:03:49] So what were some of the challenges you faced while you're paving your own lane in the field?
[00:05:34] Separate achievement from purpose
[00:06:53] The synergy of ideas
[00:10:24] The biggest positive of machine learning on society in the next two to five years.
[00:12:35] The scariest applications of machine learning in the next two to five years?
[00:14:00] The online echo chamber
[00:15:12] Big data versus thick data
[00:17:05] Is thick data like longitudinal data?
[00:19:23] As practitioners of data science and machine learning, what do you think will be some of our biggest areas of concern?
[00:21:34] The “Scott Page Canned Beets” argument
[00:24:49] What is a model and why must they be simple?
[00:26:10] What are the three classes of models?
[00:26:50] What are the seven uses of models, aka the REDCAPE?
[00:29:00] The wisdom hierarchy
[00:31:14] The importance of assumptions while constructing a model
[00:33:20] Many model thinking vs single model thinking
[00:35:53] The difficulties of modelling human behavior
[00:39:02] Identity diversity versus cognitive diversity
[00:42:42] Cognitive diversity and mental models
[00:44:43] Cognitive diversity for knowledge workers
[00:45:14] Diversity and creativity
[00:47:04] In what ways does diversity make systems more productive?
[00:48:28] Is Data science machine learning to be an art or purely a hard science?
[00:49:31] Success and creativity
[00:51:32] What's the one thing you want people to learn from your story?
[00:53:41] The lightning round