Built using transformer models, which power large language models (LLMs) like ChatGPT, the tool called Life2vec is trained on a data set pulled from the entire population of Denmark.
Life2vec is capable of predicting the future, including the lifespan of individuals, with an accuracy that exceeds state-of-the-art models, the researchers said. However, despite its predictive power, the research team said it is best used as the foundation for future work, not an end in itself.
«Even though we're using prediction to evaluate how good these models are, the tool shouldn't be used for prediction on real people,» says Tina Eliassi-Rad, a professor at Northeastern University, US. «It is a prediction model based on a specific data set of a specific population.»
By involving social scientists in the process of building this tool, the team hopes it brings a human-centered approach to AI development that doesn't lose sight of the humans amid the massive data set their tool has been trained on.
«This model offers a much more comprehensive reflection of the world as it is lived by human beings than many other models,» said Sune Lehmann, author of the study published in the journal Nature Computational Science.
At the heart of life2vec is the massive data set that the researchers used to train their model.
The researchers used that data to create long patterns of recurring life events to feed into their model, taking the transformer model approach used to train LLMs on language and adapting