Raw LLM Responses

Basil C Puglisi Thank you for sharing this. I’m glad to see independent voices thinking seriously about these questions. There are actually a lot of people and organizations wrestling with similar issues from very different angles. You have frontier AI companies like OpenAI, Anthropic, and DeepMind trying to build and align increasingly powerful systems. You have AI safety and existential risk researchers worried about long term loss of control scenarios. You also have ethics and bias researchers focused on monoculture, surveillance, and embedded power structures in current AI systems, along with governance groups trying to figure out what accountability and international oversight should look like as these systems become more capable. You probably saw Anthropic’s constitution they put out. Your paper seems most connected to the pluralistic alignment folks, particularly the question of whose values these systems are ultimately being trained to represent. Who do you see as doing work most aligned with your approach right now?