Speed Reads


Artificial intelligence is alarmingly good at predicting premature death

Artificial intelligence is getting unnervingly accurate in predicting whether people will die prematurely.

Researchers at the University of Nottingham are using artificial intelligence to help foretell a human's last days. Scientists used several AI models to analyze open-access health data and predict an individual's risk of dying young, writes LiveScience. The most advanced AI model correctly identified the risk for 76 percent of patients who died prematurely, while more traditional algorithms only identified about 44 percent of patients who were at risk of a shortened life expectancy.

"We have taken a major step forward in this field by developing a unique and holistic approach to predicting a person's risk of premature death by machine-learning," said lead study author, Dr. Stephen Weng in a statement. "This uses computers to build new risk prediction models that take into account a wide range of demographic, biometric, clinical and lifestyle factors for each individual assessed, even their dietary consumption of fruit, vegetables and meat per day."

Nearly 500,000 people in the U.K. submitted health data between 2006 to 2016. About 14,000 of those participants died prematurely, mostly from cancer, respiratory issues and heart disease. Weng said the AI models were "significantly more accurate" in predicting those early deaths. Read more at LiveScience.