OpenAI Unveils New AI Model for ChatGPT
OpenAI, the developer of the conversational AI chatbot ChatGPT, has introduced its new model "GPT-4o," which can perform real-time reasoning over text, audio, and images. According to OpenAI, GPT-4o marks a step towards more natural human-computer interaction, producing outputs from inputs consisting of any combination of text, audio, and images.
The announcement highlights that the response time of the model to audio inputs is similar to human reaction times in conversations, emphasizing GPT-4o's superior understanding of images and audio compared to existing models.
It's noted that GPT-4o is the first model trained end-to-end on a single architecture for text, images, and audio, processing all inputs and outputs through the same neural network.
Furthermore, OpenAI announced the release of a new ChatGPT desktop application for Mac operating systems, designed to integrate seamlessly with computer tasks for both free and paid users.
During the unveiling event, OpenAI mentioned that the model would be available for free due to its more cost-effective nature compared to previous models, while paid users of GPT-4o would have access to increased capacity.
The new model will be rolled out in ChatGPT in the coming weeks.
Source: Reuters
Comments
Attention!
Sending all kinds of financial, legal, criminal, administrative responsibility content arising from illegal, threatening, disturbing, insulting and abusive, humiliating, humiliating, vulgar, obscene, immoral, damaging personal rights or similar content. It belongs to the Member / Members.