[Dec. 18, 2023: JD Shavit, The Brighter Side of News]
Researchers harnessed the power of artificial intelligence to predict human life events, even going so far as to estimate the time of death. (CREDIT: Creative Commons)
In a groundbreaking research project, a collaboration between DTU (Technical University of Denmark), the University of Copenhagen, ITU (IT University of Copenhagen), and Northeastern University in the United States has harnessed the power of artificial intelligence to predict human life events, even going so far as to estimate the time of death.
This cutting-edge study, published in Nature Computational Science under the title "Using Sequences of Life-events to Predict Human Lives," has the potential to reshape the way we perceive human existence and our ability to anticipate the future.
The key to this revolutionary development lies in the utilization of transformer models, a type of artificial intelligence architecture akin to OpenAI's ChatGPT, which has been predominantly employed for processing language.
These transformer models, when trained on vast volumes of data encompassing various aspects of people's lives, demonstrate an astonishing capability to systematically organize this data and, remarkably, predict what may unfold in an individual's life, including the profoundly profound task of estimating the time of their eventual demise.
Life2vec: Unveiling the Predictive Power
At the core of this transformative endeavor is the development of a novel model known as life2vec. This model, following an initial training phase where it acquires an intimate understanding of the patterns within the data, showcases its prowess by outperforming other advanced neural networks. It remarkably provides predictions that extend to personality traits and, strikingly, the elusive time of death, all with an exceptional degree of accuracy.
Sune Lehmann, a professor at DTU and the first author of the article, expressed the research team's perspective, stating, "We used the model to address the fundamental question: to what extent can we predict events in your future based on conditions and events in your past? Scientifically, what is exciting for us is not so much the prediction itself, but the aspects of data that enable the model to provide such precise answers."
Predicting the Inevitable: Time of Death
One of the most captivating aspects of Life2vec's predictions is the ability to foresee the time of an individual's death. The model answers inquiries such as 'death within four years' with startling accuracy. As the researchers dissect the model's responses, they find that the results harmonize with established findings in the realm of social sciences. For instance, the model confirms that individuals in leadership positions or with high incomes are more likely to survive, while factors such as gender, skill level, or mental health diagnoses are associated with a heightened risk of mortality.
Life2vec employs an ingenious technique of encoding data into a vast system of vectors, a mathematical structure that meticulously organizes various aspects of an individual's life. Birth, education, income, housing, and health are all assigned their respective positions within this intricate temporal framework.
Sune Lehmann elaborates on this approach, "What's exciting is to consider human life as a long sequence of events, similar to how a sentence in a language consists of a series of words. This is usually the type of task for which transformer models in AI are used, but in our experiments, we use them to analyze what we call life sequences, i.e., events that have happened in human life."
Representation of life-sequences conditioned on mortality predictions. (CREDIT: Nature Communications)
While this groundbreaking research ushers in a new era of possibilities, it simultaneously raises profound ethical questions that demand thorough examination. Chief among these concerns is the protection of sensitive data, safeguarding individual privacy, and addressing the specter of bias within the data. These ethical dilemmas must be comprehensively understood and navigated before the model can be deployed for practical applications, such as assessing an individual's risk of contracting a disease or predicting preventable life events.
Sune Lehmann emphasizes the importance of engaging in a thoughtful dialogue on these issues, stating, "The model opens up important positive and negative perspectives to discuss and address politically. Similar technologies for predicting life events and human behavior are already used today inside tech companies that, for example, track our behavior on social networks, profile us extremely accurately, and use these profiles to predict our behavior and influence us. This discussion needs to be part of the democratic conversation so that we consider where technology is taking us and whether this is a development we want."
A schematic individual-level data representation for the life2vec model. (A) We organize socioeconomic and health data from the Danish national registers from 1st January 2008 until 31st December 2015 into a single chronologically ordered life sequence. (CREDIT: G. Savcisens et al.)
In the wake of this transformative research, the scientists involved propose an exciting avenue for further exploration. They envision expanding the model's capabilities by incorporating additional types of information, such as textual and visual data, along with insights into individuals' social connections. This holistic approach to data analysis has the potential to usher in an entirely new era of collaboration between the fields of social and health sciences.
The power to predict the future is a double-edged sword, one that holds both immense promise and profound ethical challenges. It is in our collective hands to ensure that this newfound knowledge is used for the betterment of society, fostering informed decisions and shaping a future where the boundaries of what is possible continue to expand.
For more science and technology stories check out our New Innovations section at The Brighter Side of News.
Note: Materials provided above by The Brighter Side of News. Content may be edited for style and length.
Like these kind of feel good stories? Get the Brighter Side of News' newsletter.