OpenAI conducted a survey of over 1,000 people from around the world to gather input on how artificial intelligence (AI) should behave. The results were compared to OpenAI's Model Spec, which outlines the guidelines for their AI models. The goal is to align AI defaults with diverse human values and perspectives through a process called collective alignment. --- Why it matters: This matters because it shows that large language model developers like OpenAI are actively seeking input from the public on how their technology should be designed to reflect human values, which can have significant implications for the development of trustworthy and responsible AI systems. Source: https://openai.com/index/collective-alignment-aug-2025-updates