The "Life2Vec" tool uses an operating model similar to the "GPT Chat" model, but instead of processing textual data, it analyzes life stages such as birth, school, or even working hours (Shutterstock)

Researchers in Denmark developed an artificial intelligence tool they called the “Death Calculator,” and they used the data of millions of people to help predict the stages of an individual’s life until death, in a step through which they hope to raise awareness of the power and dangers of this technology, according to Agence France-Presse.

Sonny Lehmann, a professor at the Technical University of Denmark, and one of the authors of the study published by the National Computer Science magazine, explained to Agence France-Presse that the tool is “a very general framework for predicting human life, and it can predict anything if it has training data.”

He pointed out that the possibilities are open and endless, as the tool “is able to predict health outcomes. It can predict fertility or obesity, or perhaps who will get cancer or not. It can predict whether a person will make a lot of money.”

In practice, Life2vec uses an operating model similar to that of GBT Chat, but instead of processing textual data, it analyzes life stages such as birth, school, or even working hours.

“Life is, in a sense, just a series of events: people are born, go to the pediatrician, go to school, move from house to house, get married, and so on,” the study noted.

She added, "Here we take advantage of this similarity to adapt innovations in automated natural language processing to the requirements of studying the evolution of human life and the possibility of predicting it based on a detailed sequence of events."

The “Death Calculator” analyzes the data of millions of people using artificial intelligence to help predict the stages of an individual’s life until death (Shutterstock)

Data for 6 million Danes

The tool is based on non-nominal data from about 6 million Danes collected by the National Statistical Institute.

Sequence analysis allows prediction of the remaining stages until the end of life. Regarding death, the algorithm affects 78% of cases, and in migration cases 73%.

“With a very small group of people between the ages of 35 and 65, we are trying to predict, based on an 8-year period from 2008 to 2016, whether a person will die in the next four years, up to 2020,” Lehman said. Very good, better than any other algorithm."

This age group, in which deaths are usually low, allows us to verify the reliability of the program, according to the researchers.

But the tool is not yet ready for use by the general public, because it still contains a vulnerability. “At the moment it is just a research project exploring the field of possibilities, and we do not know if it treats everyone equally.”

Scientific parallel weight

In the university professor’s view, the project represents a scientific alternative parallel to the algorithms created by the major technology companies known as “Gafam,” i.e. “Google,” “Apple,” “Facebook,” “Amazon,” and “Microsoft.”

He said, "These companies can also build models of such a tool, but they do not announce it or talk about it." The researcher noted that these companies merely use these algorithms to make the public “buy more products.”

He stressed the importance of "having an available and open parallel alternative that demonstrates what can be done with data of this type."

Data ethics expert Pernille Tranberg noted that algorithms of this type are originally used in the insurance field.

She added, "We have certainly been placed in groups, and this can be used against us, to the point of forcing us, for example, to buy an insurance policy at a higher price, and preventing us from obtaining a loan in the bank or government care, because we will die in any case."

But this does not apply to the research project, as it is not intended for individual use, given that the identity of its sources remains confidential.

She reassured that “any examples of personal data leakage” had not previously been recorded with the National Statistics Institute, and “the data are not on the names of individuals,” but she warned that “everything is accelerating” due to the development of artificial intelligence.

Commenting on some people's attempt to exploit the idea for commercial purposes, Tranberg said, "On the Internet, we already see prediction watches that indicate the age a person will reach, and some of them do not have any credibility."

Source: French