Generalisation of Total Uncertainty in AI: A Theoretical Study