Notes for AKT-140303/0:48:28

After the end of the recording we had a further discussion of the fact that "not all invariant tensors in ${\mathfrak g}^{\otimes n}$ arises from a $W_{\mathfrak g}$-like construction". See also Haviv's note, where a symmetric invariant tensor in ${\mathfrak g}^3$ is described for ${\mathfrak g}=sl(n)$, $n\geq 3$. Yet it is known that every "diagrammatic" element of ${\mathfrak g}^{\otimes 3}$ is anti-symmetric.