The Technical University of Denmark (DTU) has developed an AI-based death predictor that claims high accuracy in forecasting people’s lifespans. The AI Life2vec system, modeled after ChatGPT, makes predictions based on personal information such as health, education, occupation, and income. The model improves its accuracy by using data from Denmark’s population. The death predictor achieves a 78 percent accuracy rate after analyzing health and labor market data from 2008 to 2020 involving 6 million people.
“We use the technology behind ChatGPT (something called transformer models) to analyze human lives by representing each person as the sequence of events that happen in their life,” Sune Lehmann, lead author of the December 2023 study “Using sequence of life events to predict human lives,” told The New York Post.
“We use the fact that, in a certain sense, human lives share a similarity with language,” explained Mr Lehmann. “Just like words follow each other in sentences, events follow each other in human lives.”
“Mortality prediction is an often-used task within statistical modeling that is closely related to other health prediction tasks and therefore requires life2vec to model the progression of individual health sequences as well as labor history to predict the correct outcome successfully,” the authors of the study wrote.
Life2vec takes a different approach than ChatGPT, which focuses on crafting creative text or navigating professional hurdles. This AI delves into people’s personal histories, analyzing details such as health, education, career, and income. Life2vec hopes to predict more than just career success or fashion choices but something far more profound: the very outcome of a person’s life. It’s like holding a mirror to the past, seeking to glimpse the future it reflects.
