OpenAI reinstates ChatGPT's model picker following user backlash against its unified GPT-5 approach. The new "Auto," "Fast," and "Thinking" modes accompany restored legacy models, including GPT-4o, highlighting users' attachment to specific AI personalities. CEO Sam Altman admits the company underestimated these connections, stating, "we really just need more per-user customization of model personality." With a limit of 3,000 messages per week for the "Thinking" mode, OpenAI faces challenges in balancing performance and user preferences.