Their work & the questions they were asked at the end made me think of the extreme complexities of social systems / philosophy of violence research and the abstraction budgeting that any researcher in this space likely had to grapple with.
I feel like research on this sort of topic tends to fall into one of two bins. Philosophers, writers, artists etc focus on extremely nuanced, conceptually detailed, rich, insightful perspectives rooted in specific cultural contexts and experiences—but at the cost of limited or unclear generalization—while researchers like Erica Chenoweth take a more crude yet manageable abstraction focus, with a significantly broader scope for generalization across times, regions and contexts. But I rarely see approaches that can do both.
I imagine part of this bias-variance trade-off exists due to limited data. But I wonder if this gap also exists because, historically, a fixed amount of a single human researcher’s effort can focus on either depth or breadth, but not both simultaneously (at least not very compellingly). Perhaps AI systems could help us do new forms of research that is both deep and broad on these sorts of topics in the future on a much larger set of topics than would be possible just relying on the cooperation of individual human experts.
Like I’m thinking of something that combines an Erica Chenoweth/Our World in Data approach to understanding civil resistance with the nuance of authors like Noam Chomsky or John Paul Lederach.
I feel like research on this sort of topic tends to fall into one of two bins. Philosophers, writers, artists etc focus on extremely nuanced, conceptually detailed, rich, insightful perspectives rooted in specific cultural contexts and experiences—but at the cost of limited or unclear generalization—while researchers like Erica Chenoweth take a more crude yet manageable abstraction focus, with a significantly broader scope for generalization across times, regions and contexts. But I rarely see approaches that can do both.
I imagine part of this bias-variance trade-off exists due to limited data. But I wonder if this gap also exists because, historically, a fixed amount of a single human researcher’s effort can focus on either depth or breadth, but not both simultaneously (at least not very compellingly). Perhaps AI systems could help us do new forms of research that is both deep and broad on these sorts of topics in the future on a much larger set of topics than would be possible just relying on the cooperation of individual human experts.
Like I’m thinking of something that combines an Erica Chenoweth/Our World in Data approach to understanding civil resistance with the nuance of authors like Noam Chomsky or John Paul Lederach.