Big Policy Canvas Research Roadmap

[...] * Despite the recent dramatic boost of inference methods, they still crucially rely on the exploitation of prior knowledge and the problem of how those systems could handle unanticipated knowledge remains a great challenge.