Martin Vlach comments on Legible vs. Illegible AI Safety Problems

[Martin Vlach]

Pretty brilliant and IMHO correct observations for counter-arguments, appreciated!