Stephen McAleese comments on Prospects for Alignment Automation: Interpretability Case Study