“Scott Alexander suggests that effective AI safety regulation should focus on transparency measures, such as mandating strong whistleblower protections for employees at companies like OpenAI and Anthropic.”