The field of convergent technologies is advancing rapidly. So much so that in 2024 we may gain unprecedented insights, leading to AI-informed social modeling. Indeed, there is a convergence between social science models and technology’s ability to analyze and predict human behavior.
For example, AI-enhanced digital twins could tackle societal challenges such as net zero emissions and circularity. These issues will be transformed through hyperpersonalization powered by predictive behavioral models based on observational data from real people rather than potentially biased theoretical models.
From behavioral prediction to social modeling
The idea of a convergence between these two areas is not new. Several technologies: image recognition, behavior analysis, modeling based on behavioral sciences and psychology, are already in use.
Access to large-scale behavioral data and advanced analytics will enable a better understanding of collective dynamics, networks and complex social models. These will in turn lead to new models to uncover the precise mechanisms behind herd behavior, for example in financial markets, the dissemination of false information and the adoption of societal norms. Representing these mechanisms as digital twins for the development of potential interventions can provide the information needed for policy development.
Modeling and computational social science will simulate societies in new detail. This will make it possible to test theories and policies more thoroughly. For example, analyzing the dynamics of cities and the needs of residents could inform the planning of housing, mobility and leisure based on local contexts. New teaching models could also lead to personalized teaching, enhanced by AI – more effective for different types of learning.
The era of digital repetition
If we are not there yet, this new era is looming. The objective is to provide realistic simulations to support activity planning. This new technology combines the behavioral economics model “Prospect Theory” and AI to infer the behavior of people in the real world. It replicates human biases, such as our tendency to overestimate losses and underestimate potential gains, as well as situational factors that influence behavior, such as weather conditions.
We could thus study the impact of the use of this or that means of transport on: the traffic plan, the implementation of this or that financial advantage, the environment, etc. and more broadly transport policy.
Better behavior prediction and social modeling
Further rapid advances in the areas of behavior prediction are expected to improve quality of life and create a safe and sustainable society.
Crime prevention is one of many promising developments underway. In Japan, field trials are underway to help people avoid being scammed by phone calls by analyzing their vital data. The technology can estimate whether people are being deceived based on fluctuations in feelings of anxiety from vital data, such as breathing and heart rate.
This approach also carries risks. Privacy must be preserved and bias eliminated. However, current social modeling is perhaps even more likely to reach erroneous conclusions due to small data samples and the risk of researcher bias. AI-derived social modeling has the potential to improve our predictions about human and social behavior while improving and streamlining the operation of public services.