The post critiques OpenAI's ChatGPT memory functionality, highlighting a significant UX issue where the system retains personal information but fails to comply with explicit user instructions. The author argues that this inconsistency raises trust concerns and advocates for better transparency and accountability in AI memory operations.