While artificial intelligence (AI) is still in its early stages, its rapid evolution commands admiration. This technology is making its way into various spheres, including the education sector with initiatives such as MIA Secondary, intended to familiarize French students with AI. However, certain algorithms such as life2vec aim to go much further, claiming to be able to predict the determining events of human existence thanks to the in-depth examination of health data. This extends to the omen of the possible date of a person\u2019s death.<\/p>\n\n\n
Life2vec, resulting from the concerted work of multiple academic institutions, draws its predictions from a range of personal data from a cohort of 6 million individuals over a period extending from 2008 to 2020. This data covers various aspects such as as socio-economic status, educational level and medical records. This system has proven its ability to not only identify the behavioral characteristics of a subject but also to sense the time period leading to their death.<\/p>\n\n\n
Researcher Sune Lehmann J\u00f8rgensen explains: \u00ab\u00a0Our model was built to explore the extent to which future moments can be conjectured based on previous incidents and circumstances. From a research point of view, what we fascinates, it is less the prediction per se than the elements of information which authorize these precise forecasts.<\/p>\n\n\n
This state-of-the-art system can therefore estimate with surprising accuracy the probability of a person dying within a specific time range. Life2vec reveals that various variables affect mortality and life expectancy, an expected statement but one that AI corroborates. The designers nevertheless admit that their creation raises important ethical issues, particularly around the protection of hypersensitive personal data, respect for privacy and the existence of potential algorithmic bias.<\/p>\n\n\n
Before life2vec can be used to anticipate moments as crucial as illness or death, it will need to overcome these ethical hurdles. These efforts tend to prevent the first event and postpone the second. According to Sune Lehmann J\u00f8rgensen, this kind of careful analysis is already implemented by social networking platforms to \u00ab\u00a0finely profile us, predict our actions and exert influence over us.\u00a0\u00bb He suggests that there is a natural evolution here.<\/p>\n\n\n
The next phase for life2vec will include the integration of other kinds of data, such as those on social relationships, to further refine predictions and forge stronger links between the social and medical sciences.<\/p>\n\n\n
Keep an eye on the latest AI news by following Google News<\/a> and the chain WhatsApp<\/a> Digital.<\/p>\n\n\n