Author: Huxley
© Huxley — альманах о философии, бизнесе, искусстве и науке
5 minutes for reading

EVENT AS A WORD: Scientists have learned to foresee deaths

EVENT AS A WORD: Scientists have learned to foresee deaths
Share material
Evelyn de Morgan. Cassandra, 1898 / Artwork:


The story of the Trojan priestess Cassandra, described by Homer, shows that predicting the future is a dangerous and ungrateful endeavor. But that does not make it any less fascinating for mankind. Attempts to predict fate, and on a strictly scientific basis, are being made today. No, we are not talking about a «time machine» and quantum paradoxes, but about quite prosaic in our time technology big data. Using it, Danish scientists can now guess no worse than gypsies once did by hand.


Prediction is very difficult, especially if it’s about the future


Niels Bohr




The journal Nature Computational Science published an article titled «Using sequences of life-events to predict human lives», in which a Danish-American group of scientists described their experiment to predict fate.

The start of the project, which was led by Sune Lehmann from the University of Copenhagen, began with the collection of biographical data of all residents of Denmark without exception on January 1, 2008. The information was collected from public sources and government databases until December 31, 2015.




The scale of the work done is astounding! Everything that could be learned about the lives of 6 million people from Denmark was collected over 8 years. Everything was taken into account: changes in health status, doctor visits, marital status, change of residence, education, career dynamics, earnings, spending, etc.

It is good that this information was anonymized, otherwise the scientific research could be equated to another totalitarian version of Orwell’s «Big Brother». Would you agree that it is not very pleasant to realize that someone is trying to control not only your present but also the future?




Next, the authors of the study were engaged in analyzing the data obtained. As a basic model for their method, they turned to linguistics, from which they borrowed something very similar to the semantic analysis of text.

That is, the system extracted «semantic clusters» from the database, on the basis of which a pattern of the statistical probability of early death in the age category from 35 to 65 years was deduced. The accuracy of the forecast reached an unprecedented level of 40–42%!

This is all the more surprising because no prediction system, including those based on the capabilities of artificial intelligence, has been able to give such accuracy. Even for the now fashionable neural networks, this figure does not exceed 30%. 


By joining the Huxley friends club, you support philosophy, science and art




The obvious correlation between the principles of organizing the «text of language» and the «text of fate» brings to mind the postmodern intuition of philosopher Jacques Derrida. For him, the textuality of the world was total, the whole world was presented to him as a text. Every object or event was also a text.

According to Derrida, text is the only possible model of reality as such. What then to speak about the «text of destiny» of an individual human being, a being endowed with linguistic, i.e., «textual», consciousness and thinking?

In this regard, the effectiveness of semantic analysis technologies for predicting fate seems far from accidental. Philosophers, in relation to the history of science, become something like a «collective Cassandra» — nobody believes in their «abstractions» for the time being either.




It turns out that a person’s biography is made up of events about as much as text is made up of words. An article in Nature Computational Science states that the analysis of biographical facts to predict the future can be used on a quite scientific basis, devoid of all the mysticism characteristic of the «dark» past.

In any text, especially such a complex one as human life, there is a considerable number of hidden, very intricate, and non-obvious cause-and-effect relationships. However, analysis technologies explicitly designed for data with built-in «memory» have learned to recognize them in much the same way as in language texts. 




Since, in this model, an event is no different from a word, the authors of the study named their program Life2vec by analogy with Word2vec. However, another aspect is not less interesting: in addition to predictions, the program perfectly copes with making a psychological portrait and describing the personal characteristics of a person.

So it turns out that personality and destiny are interdependent in some way, and a person’s personality is as «text-like» as his biography? Is it possible that they are one and the same text, only written in different «languages»?




The study of the Danes can be made very unusual for «classical» science conclusions, which, however, long ago realized the ancient philosophical and religious systems. But if we abstract from the «general cultural» component, the study, even in a strictly applied sense, is quite promising.

After all, thanks to it, we can learn more about the factors of early mortality, of which we had not the slightest idea before. Thus, Bulgakov’s Voland’s statement that «a brick does not fall on anyone’s head by accident» turns out to be not only accurate but also acquires strictly scientific content.


Original research: Using sequences of life-events to predict human lives


When copying materials, please place an active link to
By joining the Huxley friends club, you support philosophy, science and art
Share material

Spelling error report

The following text will be sent to our editors: