Technical University of Denmark (DTU), the University of Copenhagen, the IT University of Copenhagen (ITU), and Northeastern University in the U.S. have come up with a new yet slightly scary AI model called AI Death Calculator - Life2vec.
“Predicting when you’ll die and estimating your finances as that time approaches…,” reads the descriptor on Life2vec’s official website.
The official definition reads, “The AI Death Calculator is a digital tool that employs artificial intelligence (AI) technology to predict your life expectancy or estimate the date of your death based on various lifestyle and health factors.”
Trained on a data set containing an in-depth registry of every citizen in the Scandinavian country. (around 6 million people), it claims to use artificial intelligence to predict an individual's likelihood of dying within the next four years.
The study represents human lives in a format resembling language structure, using natural language processing (NLP) techniques to analyze and predict life trajectories. It includes detailed information on health, education, occupation, income, address, and working hours. The researchers create embeddings of life events in a unified vector space, demonstrating its robustness and structure.
The models excel in predicting various outcomes, surpassing current models significantly. By interpreting deep learning models, the researchers explore factors contributing to predictions, offering insights into potential mechanisms influencing life outcomes and personalized interventions.
Despite the sensationalism surrounding the AI tool's portrayal as a "doom calculator," the research delves deeper into the merger of AI and social science, exploring correlations and patterns in human life events. The project, which tokenized data from 2008 to 2015, aimed to understand high-dimensional event sequences in people's lives and raise questions about societal patterns.
The researchers converted various life data into numeric tokens, allowing the AI to identify patterns and correlations that may not be apparent to humans. Using meticulously organized data from Statistics Denmark, the study mapped out the seemingly chaotic aspects of a modern Danish person's life.
However, the study's primary focus was not on predicting death but on uncovering intricate details about human behavior and societal trends. It suggests that countries with robust AI and rich datasets, such as Denmark, could gain significant advantages in efficiently managing budgets and directing social services.