OpenAI applied for GPT-6 and GPT-7 trademarks in China, saying that GPT-5 was already in training.

Thanks to CTOnews.com netizens Colorful M, West window past, Xiao Xing _ 14, Love sharing, Cola Kele, Baa Yang's clues delivery! CTOnews.com December 3 news, recently, ou'ai Operation Co., Ltd. (OPENAI OPCO,LLC) applied for a number of "GPT-6" and "GPT-7" trademarks, internationally classified as scientific instruments and website services, and the current trademark status is waiting for substantive examination.

At present, ChatGPT 4.0 is the latest generation of artificial intelligence chat robot model of OpenAI company. ChatGPT 4. 0 is based on the Transformer architecture of self-attention mechanism. This is a model based on deep learning. Through a large amount of data training, it can capture complex patterns in the text and generate new and coherent text. Its working principle is to predict the next word according to the previous article, and then iteratively generate the whole text. ChatGPT 4.0 has the advantage of a large number of parameters, can handle many types of text tasks, and performs well in many fields.

CTOnews.com noted that OpenAI CEO Sam Altman had previously revealed that OpenAI was actively developing the next generation of AI model GPT-5. Although the release schedule is uncertain, he made it clear that more data is needed to train the model, so the company is calling for large-scale data sets, especially long writing or dialogue data that are not yet publicly and easily available on the Internet.