The 'pause on large language model training' proposal is not implementable or obviously beneficial because it would require stopping all competing AI development simultaneously across all nations and actors, which cannot be verified or enforced, and because alternative AI approaches (beyond LLMs) that could be deployed for harmful purposes are not addressed