OpenAI updates GPT-5 with new personality and modes
CEO Sam Altman highlighted that reasoning-focused model use is growing but still limited, prompting updates to GPT-5 for better user control.
OpenAI has introduced updates to its GPT-5 model following user feedback. CEO Sam Altman announced that users can now choose between Auto, Fast, and Thinking modes, along with an updated personality for the AI.
The changes aim to enhance user experience by providing greater control over the model’s behaviour. Altman noted that while more users work with reasoning-focused models, they still represent a relatively small portion of the total user base.
The update reflects OpenAI’s ongoing commitment to tailoring AI interactions based on user preferences and feedback, ensuring the flagship model remains adaptable and responsive to diverse needs.
GPT-5 faced a rocky launch as users found it surprisingly less capable than GPT-4o, due to a malfunctioning real-time router that misrouted queries. Sam Altman acknowledged the issue, restoring GPT-4o access and doubling rate limits for Plus subscribers.
The episode has also sparked debate in the AI community about balancing innovation with emotional resonance, as some users note GPT-5’s colder tone despite its safer, more aligned behaviour.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
