Development of GPT – 5 and Progress in AGI

​OpenAI recently unveiled progress on GPT-5 development, emphasizing the critical balance between reasoning capabilities and conversational fluency.​​ Core researcher Michelle Pokrass highlighted that GPT-5 aims to train a model capable of dynamically adapting its thought processes and dialogue style based on contextual nuances. Meanwhile, Chief Research Officer Mark Chen underscored that AGI extends beyond ChatGPT, encompassing a broader ecosystem of intelligent agents. The team has achieved breakthroughs in fine-tuning methodologies, with Reinforcement Fine-Tuning (RFT) demonstrating exceptional performance across diverse domains. OpenAI is also committed to open-sourcing reasoning models, striving to harmonize commercialization with scientific exploration. Expressing optimism about AI’s future, the team encouraged newcomers to maintain curiosity as a guiding principle.