Groundbreaking use of Danish AI to predict life outcomes

Scientists in Denmark are making use of artificial intelligence and data from millions of people to forecast the stages of a person’s life up to the end, aiming to show the capabilities and risks of this technology.

The creators of life2vec are not fixated on morbidity, but more so on exploring the patterns and relationships that deep-learning programs can reveal to predict health or social “life events.”

Sune Lehmann, a professor at the Technical University of Denmark, and co-author of a study published in the journal Nature Computational Science, explained that the algorithm has the potential to predict a wide range of outcomes, such as health issues, fertility, and financial success, by analyzing variables like birth, education, and work schedules.

The model is based on anonymized data from roughly six million Danes, gathered by Statistics Denmark, and has shockingly shown high accuracy in predicting events like mortality and relocation.

Lehmann stressed that the software is currently restricted to research purposes and not available online. He talked about the importance of transparency in developing predictive algorithms during the middle of some deep-rooted concerns about privacy and potential discrimination.

Pernille Tranberg, a Danish data ethics expert, raised flags about the potential misuse of such algorithms by businesses, including insurance companies, to discriminate against people based on predicted outcomes. While the capability to predict life events using AI is groundbreaking, the researchers acknowledge the ethical considerations that must be addressed before this technology can be widely used beyond research settings.