After 150 years of the Periodic Table of Elements, one would imagine that the structures of common elements are well known. While this is true for most elements, there are some that has defied understanding. The certainty with which the allotropes of a familiar element carbon, viz diamond and graphite, can be ascertained and understood, makes us feel that this should be true for the neighboring element boron as well. However, nature continues to play dice: structures of allotropes of boron throw up unexpected problems. It appears as though each experiment leads to a different observation. These are not necessarily due to the limitations of the measurement, but inherent in the nature of the electronic structure which results in “one” set of position of atoms in one experiment and “another” set of positions in another experiment. This is true of both the 3D and the more recent 2D allotropes of boron. Have we reached the limit of understanding, or do we have to redefine “understanding”. Does reproduction of experiment by artificial intelligence and neural networks based algorithms constitute understanding?