Researchers in Denmark have developed an AI, life2vec, capable of predicting various life events, including the end of life, using data from millions, highlighting the technology’s potential and risks.
Main Points:
Life2vec, a Danish AI algorithm, uses deep-learning to predict health and social outcomes from birth to death, based on extensive data analysis.
The algorithm, leveraging data from about six million Danes, has shown a high accuracy rate in predicting mortality and mobility, amongst other life events.
The project serves as a public, scientific counterpoint to similar, less transparent efforts by large tech companies, underlining the importance of ethical considerations and potential misuse in insurance and banking.
Summary:
In Denmark, researchers have harnessed the power of artificial intelligence to create life2vec, an innovative algorithm that aims to anticipate the stages of an individual’s life, from birth to death, by analyzing patterns in data from millions of people. The project, spearheaded by Sune Lehmann of the Technical University of Denmark, not only seeks to demonstrate the expansive predictive capabilities of AI in determining health outcomes and social events but also to raise awareness about the ethical implications and potential misuse of such technology.
Life2vec operates by analyzing detailed sequences of life events, similar to how language-processing algorithms work, but focusing on variables like education, social benefits, and employment patterns. With data from around six million Danes, it accurately predicts mortality in 78% of cases and relocation events in 73% of instances. However, despite its promise, the tool remains within the realm of research, with access restricted to prevent misuse and ensure privacy.
The development of life2vec presents a critical, open discussion point against the backdrop of similar, opaque AI endeavors by large tech corporations, primarily aimed at commercial gains. This project underscores the need for transparency and ethical guidelines in AI research and application, particularly as similar algorithms find use in sectors like insurance and banking, where they could potentially lead to discriminatory practices. Pernille Tranberg, a Danish data ethics expert, highlights the importance of public awareness and ethical considerations in the deployment of predictive algorithms, which could otherwise impact individuals’ access to services based on algorithmically determined life outcomes.
