Scientists have unveiled an artificial intelligence (AI) algorithm, dubbed as the "doom calculator," that can predict individuals' chances of dying within four years with an impressive accuracy of over 75%.
The breakthrough AI algorithm, known as "life2vec," operates similarly to ChatGPT but without direct user interaction. The researchers, hailing from Denmark and the U.S., published their pioneering project in the Nature Computational Science online journal.
Life2vec analyzed data from over 6 million individuals in Denmark, covering factors such as age, health, education, employment, income, and life events. The Danish government supplied this comprehensive dataset, which formed the basis for the AI's training, according to USA Today.
How Does the Doom Calculator Work?
Life2vec was trained to process information about people's lives, presented in the form of sentences. For instance, it learned from sentences like "In September 2012, Francisco received 20,000 Danish kroner as a guard at a castle in Elsinore" or "During her third year at secondary boarding school, Hermione followed five elective classes." As the AI progressed, it gained the ability to construct "individual human life trajectories."
Lead author Sune Lehmann, a professor of networks and complexity science at the Technical University of Denmark, metaphorically described human life as a "giant long sentence," encapsulating numerous events.
The AI's predictive accuracy reached an impressive 78%, correctly identifying individuals who had passed away by 2020. Crucially, none of the study participants were informed about their predicted death dates, according to The Science Times.
Factors associated with earlier deaths included mental health diagnoses, male gender, and skilled professions. Conversely, leadership roles at work and higher income were correlated with longer lifespans. The AI exhibited versatility in predicting various aspects, from personalities to decisions about international relocations.
A Breakthrough Tech But Too Risky
Despite its potential, the doom calculator is not yet fit for public use, and the associated data remains confidential to safeguard privacy. The researchers are actively exploring ways to share results more openly, prioritizing individual privacy.
Collaborator Tina Eliassi-Rad, a computer science professor at Northeastern University in Boston, cautioned against using tools like life2vec to predict individual outcomes. She stressed their utility in tracking societal trends rather than foreseeing individual futures, acknowledging that "real people have hearts and minds."
Sune Lehmann emphasized that the AI model "should not be used by an insurance company, because the whole idea of insurance is that, by sharing the lack of knowledge of who is going to be the unlucky person struck by some incident, or death, or losing your backpack, we can kind of share this this burden," as quoted by The Sun.
He also expressed concerns that major tech companies with vast amounts of data might already be utilizing similar models to formulate predictions about individuals.
The ethical implications of AI in predicting death are substantial. Art Caplan, a bioethics professor at New York University Langone Medical Center, anticipates consumers seeking their forecasted data, anticipating challenges and conflicts over third-party access to sensitive information.
Caplan notes the potential benefits of preventing deaths but raises concerns about the algorithm's impact on removing life's unknowns, which may not necessarily be beneficial.
