Can This Doom Calculator Really Predict Your Death?

What if there was a way to predict your lifespan and financial future with astonishing accuracy?

Imagine having a tool that could show you when you will pass away and how much wealth you will accumulate in your lifetime. What about predicting your death as well? Sound unrealistic right? Well, not really!

Scientists have invented a unique “doom calculator” that uses unique algorithms to calculate your estimated lifespan and projected earnings. But, does it truly work?

Doom Calculator That Predicts Your Death

Scientists have created an algorithm named ‘life2vec’ that predicts a person’s life trajectory and lifespan based on their life story.

This model, according to a recent study, achieves an accuracy rate of approximately 78 percent. It is comparable to other algorithms used for similar predictions.

Unlike other models, the ‘life2vec’ algorithm operates interactively, resembling a chatbot, to predict future outcomes based on existing data. Developed by scientists in Denmark and the US, it was trained on extensive Danish data covering over six million individuals. It incorporates details such as income, profession, residence, injuries, and pregnancy history into its predictions.

What Was the End Result?

The scientists developed a model capable of interpreting natural language and making predictions regarding a person’s risk of premature death or their income trajectory across their life. Factors associated with earlier death include being male, having a mental health diagnosis, or working in a skilled profession. Conversely, higher income and holding leadership positions are linked to longer life expectancy.

Life2vec functions by treating each phase of a person’s life like words in a sentence, predicting future outcomes based on past information. Similar to how users prompt ChatGPT to create songs, poems, or essays, scientists can ask life2vec straightforward questions such as “death within four years?” for specific individuals. The model’s training data spans from 2008 to 2016.

What Did the Doom Calculator Predict Successfully? Did it Also Predict Death?

The algorithm accurately predicted the individuals who had passed away by 2020 in over 75% of cases based on population data. The research was published in Nature Computational Science. Lead researcher Sune Lehmann emphasized that to safeguard personal information, the system’s data is not accessible to the public or companies.

Sune Lehmann, professor of networks and complex systems at the Technical University of Denmark, stated that they are actively exploring methods to share some of the findings more openly. However, this effort requires additional research to ensure the privacy protection of the individuals involved in the study.

Legality of the Doom Calculator “life2vec”

Even when life2vec becomes publicly available, Danish privacy laws would prohibit its use in individual decision-making such as insurance policies or hiring decisions. Similar to ChatGPT and other large language models trained on extensive written data, life2vec learns from detailed life events structured as data-rich sentences.

For instance, examples include sentences like “In September 2012, Francisco received twenty thousand Danish kroner as a guard at a castle in Elsinore” or “During her third year at secondary boarding school, Hermione followed five elective classes.” Professor Sune Lehmann and his team utilized unique tokens to represent each piece of information, mapping these data points in relation to one another.

Mapping Life Categories in “life2vec” and How it Predicts Death

Life stories include a majority of human experiences, each coded with specific identifiers in life2vec: for example, a forearm fracture as S52, working in a tobacco shop as IND4726, income represented by 100 different digital tokens, and ‘postpartum hemorrhage’ as O72. Some relationships, like profession influencing income, are straightforward and intuitive.

However, what distinguishes life2vec is its ability to intricately map the complex web of factors shaping an individual’s life. This enables predictions based on millions of people and a multitude of interconnected variables.

It Can Also Predict Personalities

In addition to life outcomes, life2vec can forecast people’s personality traits. Researchers, led by Professor Sune Lehmann, trained the model to predict responses to personality test questions. These questions gauge attitudes on statements such as ‘The first thing that I always do in a new place is to make friends’ or ‘I rarely express my opinions in group meetings.’

All the data used for training were sourced from Denmark, implying that predictions may not generalize to other regions. Moreover, Lehmann emphasized that while the model raises important discussions about personal forecasting, many individuals may not wish to know such predictions, especially concerning life expectancy.

Predicting Death: Ethical Approach or Not?

Similar to existing technologies used by tech companies to predict life events and human behavior through social network tracking and accurate profiling, life2vec raises ethical concerns. These technologies utilize profiles to forecast behavior and influence individuals.

Professor Sune Lehmann emphasizes the importance of including this discussion in democratic conversations. It prompts reflection on the direction technology is leading us and whether such developments align with societal values and preferences.

Would you ever use a doom calculator?

You might also want to read: The Simpsons Predictions That Actually Came True