OpenAI is discontinuing its GPT-4o model, a move that has caused significant concern among users in China who rely on the model through various access methods. The company announced the model’s deprecation as part of its standard product lifecycle, planning to replace it with newer iterations. This decision has disrupted a large community of Chinese developers, researchers, and enthusiasts who, due to regional restrictions, often access OpenAI’s models through unofficial channels, third-party APIs, or VPNs. The user base had come to depend on GPT-4o’s specific capabilities and performance characteristics. The article details the logistical and emotional impact on these users, who now face the challenge of migrating their projects and workflows with limited official support, highlighting the broader tensions of global AI development amid geopolitical boundaries. Read the full article at: https://www.wired.com/story/openai-nuking-4o-model-china-chatgpt-fans-arent-ok/