AI can simulate humans to complete a series of project tests, so why do we still need people to do all these repetitive project tests? I think the real reason airdrops in the crypto space have become obsolete and reached their end—besides the worsening environment in the crypto world—is the rise of AI large models, AI agents, workflows, and AI programming and testing. Now that even some human jobs are gradually being replaced by AI agents, how could these repetitive project testing tasks in the crypto space stand a chance?