Death Calculator: How to predict the exact time a person will die

What would you do if there was a machine that could predict your time of death?
Will this bring panic to humans?
An international research team from the Technical University of Denmark has released a new large-scale model product called: “Life2vec”, which literally means “life direction changer”. It can directly predict a person’s mortality rate with an accuracy of up to 78% %.
This is said to be the most accurate prediction method on the market.
To predict when “someone” would die (an anonymous person chosen at random for the study), the team used data on more than 2.3 million people aged 35 to 65 over a seven-year period, from January 1, 2008, to December 31, 2015. This is the current insurance data for almost all middle-aged people in Denmark. This group was chosen for the study because mortality in this age group is more difficult to predict.
The model, called Life2vec, uses this data to infer a person’s probability of surviving in 2020, four years after 2016.

“To test how effective ‘Life2vec’ is, we selected a group of 100,000 people, half of whom survived and half of whom died,” said Dr. Sune Lehmann . The researchers knew who died after 2016, but the algorithm didn’t.
In the experiment, the team found all the data on the lives of 6 million Danes for 12 years (2008-2020): address, school, medical treatment, diagnosis, income and occupation, etc., and converted it into a large language model that can be used to train text. This model is similar to ChatGPT, which analyzes large amounts of text data to predict the next most likely word to infer the likelihood of future events. In the same way, the “Life2vec” model developed by researchers can analyze the sequence of events in an individual’s life course and predict what is most likely to happen next.
The “Life2vec ” model uses data from 2008 to 2016 for training, and data from 2016 to 2020 is used for testing. Researchers divided people aged 35-65 into two groups, half of whom died between 2016 and 2020, and the other half who survived. they letThe ” Life2vec ” model predicts who will live and who will die in 2020 for a randomly selected “someone”. Finally, at the end of 2020, it was found that the prediction accuracy reached 78%.
The “Life2vec” model predicts the survival and death outcomes of these people with an accuracy rate that is 11% higher than existing AI models and mortality statistics tables commonly used in the insurance industry.

Analyzing the results of the model’s answers, the researchers found that, all things being equal, being a skilled worker such as an engineer or being diagnosed with a mental health problem such as depression or anxiety would also lead to an earlier death and a higher mortality rate. But individuals who held managerial positions or had high incomes had higher survival rates.
Most of the deaths that the “Life2vec” model failed to predict involved unpredictable accidents or heart disease.

The AI ​​Death Calculator is currently in free public beta: https://deathcalculator.ai/. Answer a few questions to test your time of death!

“Life2vec” can not only predict your death time, but also guess your personality
This AI model that can predict the time of human death: “Life2vec” was released by a research team from the Technical University of Denmark (DTU), the University of Copenhagen, the International Telecommunication Union (ITU) and Northeastern University in the United States.
A research article titled “Using Life Event Sequences to Predict Human Lifespan” published in Nature Computational Science at the end of December this year disclosed the principles and methods of this prediction.
If you use large amounts of data about human life to train so-called “transformer models” (similar to ChatGPT) to process language, they can systematically organize the data to generate predictions about whether a person is likely to die young, or his or her lifetime earnings. .

The working principle of the “Life2vec ” model is based on a large language model (similar to the language model behind ChatGPT), which summarizes patterns and rules by analyzing the sequence of events in human life. All parts of a person’s life are given a code, for example S52 means a broken forearm, 072 means postpartum hemorrhage, POS3513 means someone is a computer systems technician, etc.
The model was trained using health data and labor market dependencies of approximately 6 million Danes (population as of February 2021 was 5.935 million). This is a figure for almost the entire Danish population. In other words, they can predict the time of death or some major health event in a lifetime for almost all Danes.
Lehmann, the first author of the article and a professor at the German Technical University, said: “We are trying to use this model to solve a basic question: How likely are we to predict future events on the basis of past conditions? In science, let’s It’s not the predictions themselves that are exciting, but the data that enables the models to provide such precise answers. ”

The ‘Life2vec ‘ model is able to predict a wide range of outcomes, from premature death rates to subtle differences in personality. The researchers say their framework allows them to uncover potential mechanisms that influence lifespan outcomes and the associated possibilities for personalized intervention.
The original design of Life2vec is not just to predict mortality, it can also infer your personality based on your life trajectory.
This research article disclosed that they selected some questions from the more authoritative models in personality testing, and then randomly found some “someones” in the database to take the personality test.
For example, people who score high on the question “I prefer working with others rather than working alone” are more likely to be social. Then the human results were compared with the model predictions, and it was found that the prediction accuracy was much higher than that of the neural network algorithm.
“For the first time, Life2vec truly stands from a personal perspective, providing a possibility to glimpse what life will look like in the future from today’s choices.”
Of course, the “Life2vec ” model still has many flaws when predicting the time of death. For example, it cannot predict how a person will die . For example, algorithms cannot predict whether a person will die in a car accident or be poisoned.
But Lyman believes that in no more than five years, with the training of large amounts of data from countries around the world, perhaps we will see a more accurate model that can predict your more accurate time of death, and even death. Way.

Do humans need to know when they will die? This is probably a bad invention! What unpredictable dangers will it bring us?
What is the role of such a large death prediction model?
Researcher Lehmann believes: “This model may one day help identify a person’s disease risk and help them take timely measures to stay healthy. It can also be applied to a wide range of health and social issues, such as predicting and early intervention of health problems, Or help governments close the gap between rich and poor . But such an application would also bring with it a host of privacy, ethics, and data security issues that need to be addressed before this model can help anyone.”

This “death prediction model” was trained based on real life data of Danish people, and the predicted results are currently only applicable to Denmark. Beyond that, most people probably don’t actually want to know when they’re going to die! Lehmann said that in order to protect the personal privacy information of Danish citizens used to train the system, the AI ​​model is not currently open to the public or insurance companies.
However, the release of “Life2vec” still triggered discussion and panic in Denmark and around the world.
The CNN article believes that this model may cause harm if abused by enterprises . In particular, insurance companies, if they use this model to predict people’s health and death time, they may refuse many people who need insurance.
“The Life2vec model should not be used by insurance companies,” Lehmann emphasized in an interview. ” The essence of insurance is to share risks and predict who will encounter unfortunate events or die, which goes against the concept of mutual insurance. ”
But people are worried that this large model for predicting the time of human death, or other large models of the same type, will one day be opened and used in the real world by some institutions.
If one day you go to a company to apply for a job, and the company enters your information into the system, thinks you are a health risk, and refuses to hire you, this may become a normal tragedy.
And this will affect people’s employment and lives.
And if one day you know the time of your death, will it affect your courage to live? The unknown is the driving force for a person to live. And knowing that you are about to die, how will you spend the rest of your life?
To some extent, this large model for predicting the time of death may be a “bad invention.”
And it is impossible to predict the disastrous consequences it may cause.

error: Content is protected !!