Neural Networks, Psychology, Digital Tribes & Agent based Modeling
Neural Networks are not the only concept data science and AI has borrowed from psychology and brain sciences. Reinforcement Learning has…
Neural Networks are not the only concept data science and AI has borrowed from psychology and brain sciences. Reinforcement Learning has deep roots in agent-based modeling in social psychology. Simulations of such environments provide valuable insights into group and system-level interactions.
Reward-Punishment is already a well-documented concept in behavioral psychology and is now used to optimize outcomes with additional compute power that modern computers can afford. I talked about that with Joe Reis’ on his show. This helps us model user behaviors across different levels of interactions on the website.
Just like social identities, people have digital identities and digital tribes, and herd behavior becomes easy to predict. Bucketing these datasets into bins with parameters elicited from classification trees is the secret sauce. Digital identities can be grouped and canvassed just like mobs. The gatekeepers like Facebook knew it all along and they could prevent it. They just chose to profit from it.
Here is a paper published by Harvard’s DASH initiative on Agent-based Modeling: A Guide for Social Psychologists