Predicting Life Events with Artificial Intelligence Language Models

A study by DTU, University of Copenhagen, ITU, and North eastern University in the US has shown that language processing models, such as ChatGPT, can be utilized to predict life events. By training transformer models with significant amounts of personal data, researchers can systematically organize the data and make predictions about future events, including estimating a person's time of death.

The latest scientific research has been published in Nature Computational Science. The article, titled "Using Sequences of Life-Events to Predict Human Lives," examines the health data and labour market participation of six million Danish citizens. The study utilizes a unique model called life2vec to analyse the data.

A complex set of data is used to create a highly accurate predictive tool. The researchers have developed a powerful tool that can predict the future, such as the lifespan of individuals, with a level of accuracy that surpasses state-of-the-art models. While this tool has extraordinary predictive capabilities, the research team suggests that it should be used as a starting point for future work, rather than a final solution in itself.

“Our framework allows researchers to identify new potential mechanisms that impact life outcomes and associated possibilities for personalized interventions,” wrote lead author Sune Lehmann, a professor at Technical University of Denmark, along with co-authors Germans Savcisens, Tina Eliassi-Rad, Lars Kai Hansen, Laust Hvas Mortensen, Lau Lilleholt, Anna Rogers, and Ingo Zettler.

Understanding life events and mortality probability with a model. This model operates by observing millions of life event sequences and building vector representations in embedding spaces. It then sorts and identifies relationships between life events, such as income, education, and health factors. These embedding spaces lay the groundwork for the model's predictions. One of the most notable predictions made by the researchers was an individual's likelihood of mortality.

By utilizing a robust transformer and vast training data, recent research has shown that AI can accurately predict human life events. This breakthrough has the potential to provide valuable insights in the fields of health research and social sciences. The validation of the research prototype is a significant step forward in this area.