[tomrod]

Would you mind talking to how black box interpretability is becoming well known? I've seen Shapley values used for feature interpretation, but not sure what else is being done.

[polkapolka]

For an accessible recent overview see: https://christophm.github.io/interpretable-ml-book/