They discuss the challenges of evaluating bias in the GPT system and their commitment to improving the system and addressing these issues, while also highlighting the pressure from click bait journalism. They also discuss the OpenAI moderation tooling for GPT and their continuous learning to improve the moderation process.