Now researchers have developed an AI that can draw conclusions about people’s life and death by using public personal information such as income, health, work, place of residence and doctor visits. Among other things, the AI can predict who is at risk of dying prematurely.
The AI searches through people’s life histories
The AI works much like the Chat GPT, which searches through and finds connections in large amounts of text. Here, it is statistics about all of Denmark’s six million inhabitants that form the basis of the AI system’s predictions. To test the program, the researchers used information from the years between 2008 and 2016, and then had the AI predict whether a person between the ages of 35 and 65 died or not in the next four years. And the result was close to the real numbers.
When the program predicted who in the statistics would be at risk of dying prematurely, it proved to be correct with almost 80 percent certainty, the study shows in Nature Computational Science.
Hope the program can save lives
– Imagine if you could use an algorithm like this to identify at an early stage those who are at risk of dying from a serious illness and put in preventive measures, says Sune Lehmann who is a professor of Social Computer Science at the Technical University of Denmark.
He believes that companies already use similar programs to manage personal data and predict people’s behaviour.
– The big tech companies already have so much data and make predictions about us. We have to start thinking about what is possible and regulate it, believes Sune Lehmann.
Play the video to hear about the ethical risks of the death calculator.