I like this critique of my post! But I also think it undervalues the practical, pragmatic amount of work AI can do right now (even including its error rate). My fear remains that thoughtful critics don't see the the full risks (& benefits) of AI because they do not see its current effectiveness.