OpenAI has introduced GPT-4o, a cutting-edge AI model designed to enhance human-computer interaction. This innovative model seamlessly accepts text, audio, and images as input, generating outputs in all three formats. Notably, GPT-4o can recognize emotions, allows interruptions during conversation, and responds almost as quickly as a human.
Mira Murati, OpenAI’s CTO, stated, “The special thing about GPT-4o is it brings GPT-4 level intelligence to everyone, including our free users. This marks a significant advance in ease of use.” During the demonstration, OpenAI showcased GPT-4o translating in real time between English and Italian, assisting a researcher with a linear equation, and guiding another executive in deep breathing exercises by simply listening to his breath.
The “o” in GPT-4o stands for “omni,” highlighting the model’s multimodal capabilities. Unlike previous versions, GPT-3.5 and GPT-4, which transcribed spoken input into text and often lost emotional nuance, GPT-4o processes all inputs and outputs through a single neural network, enhancing interaction quality.
OpenAI will roll out GPT-4o to all users, including free ChatGPT users, in the coming weeks. Additionally, a desktop version of ChatGPT for Mac users will be available for paid subscribers beginning now. This launch precedes Google I/O, where Google teased a similar AI chatbot, Gemini, emphasizing the growing competition in AI technology.