Strong and Vital No 1

Life time

Would you Like to Know the Date of your Death ? How to Predict the Fatal Day

A team of researchers at the Technical University of Denmark (DTU) has deve loped an algorithm called Life2vec , which is able to estimate a person’s risk of illness and even pre dict the probable date of his or her death, provided that no unexpected inci dents occur. The study was published in the jour nal Nature Computational Science . However, the sci entists point out that this is a research prototype that cannot fulfil any real tasks in its current state. The researchers analysed data from a national register in Denmark on events in the lives of six million Danes. To do this, they collated large amounts of data on their health - from doctor visits or emergency admissions to various diagnoses - education, income and occupation. Most of this data cove red the period 2008 to 2016. To test how good the AI tool is, a group of 100,000 people was selected, half of whom survived and half of whom died. The researchers knew which people had died after 2016, while the algorithm did not. Then they put it to the test: they had the algorithm make individual predictions about whether someone was still alive after 2016 or not. The results were impressive.

The new AI tool was cor rect in 78 per cent of cases. Among other things, the study showed that people with a high income or in a managerial position were more likely to survive, whe reas men, skilled workers or people diagnosed with a mental disorder were more likely to die. Apart from the question of how reliable such a model is, this futuristic project

naturally also raises important ethical questions regarding the use of sensitive medical data and its confidentiality. Nevertheless, this type of research opens up new and unexpected perspectives, as the team notes. Insurance companies will endeavour to stay one step ahead of cus tomers if models such as Life2vec become established, which will make it more difficult to underwrite insurance in the future. It would be problematic, for example, to take out life insurance if you know exactly what the health risks are. Ultimately, according to the critics, using artificial intel ligence to predict the time of death with high accuracy would take away the only aspect of our lives that makes them mysterious in some way.

Extending life expectancy

• Intellectually motivated : The pleasure of having an active mind, be it reading, playing cards, doing cross word puzzles or Sudoku. Curiosity, love of learning and self-teaching. • Positivity : Being grateful for life and knowing how to enjoy it despite difficulties. • Resilience : The ability to overcome adversity and even be strengthened by such experiences. • Intelligence : General cognitive ability to think abstractly, to reason, to plan, to successfully solve pro blems, to learn from them quickly. Accepting challenges, practising reactivity and problem-solving.

• Vitality: Feeling alive, alert and full of energy. Having the will to live long. Taking part in events, not standing on the sidelines. Use opportunities to be active. • Interaction : Socialising while avoiding loneliness, forming warm bonds, practising altruism. • Commitment : Responsibility with a sense of duty, perseverance, honesty. • Control : The conviction that you can influence the course of events, purposefulness, practicality. Orga nising life, using and understanding the environment, recognising and exploiting opportunities.

According to the WHO, the number of centenarians will increase tenfold by 2050 compared to 2010. A study has now analysed the characteristics of people over 100. A team of researchers led by psy chologist Dolores Merino from the Faculty of Psychology at the Complu tense University of Madrid published a study in the J ournal of Happiness Studies in November 2023, in which the behavioural characteristics of 19 centenarians were analysed. This pro vides clues as to how we can remain healthy and happy even in old age:

14

STRONG and VITAL No. 1 - 2024