The now hot ChatGPT is derived from OpenAI’s large-scale language model GPT-3.5. The market has been rumored that the next generation of GPT-4 will be released soon. Andreas Braun, technical director of Microsoft Germany, confirmed that GPT-4 will be released next week. He also revealed that it supports multimodal models and has more innovative functions, such as generating videos. It even describes GPT-4 as a game changer.
Andreas Braun revealed at Microsoft Germany’s hybrid event “KI im Fokus” (English: AI in focus) that Microsoft will announce GPT-4 next week. It is announced that GPT-4 has a multi-modal model, which can understand text and translate it into different languages at the same time, such as inputting Chinese questions to get English answers. You can even convert text commands into pictures, music or videos.
A further example is the customer service center, which can use GPT-4 to automatically turn the phone conversations between employees and customers into text records, saving a lot of time and manpower to sort out the content afterwards.
Andreas Braun did not further explain the details of the publication, but Microsoft has issued an invitation to an online presentation “The Future of Work: Reinventing Productivity with AI” next Thursday. The time is Thursday (16th) at 8:00 a.m. US West Coast time (11:00 p.m. Hong Kong time on the same day). The presentation was moderated by Satya Nadella, Chief Executive Officer, and Jared Spataro, Vice President, Modern Work and Business Applications. GPT is not Microsoft’s technology, but Microsoft has invested $10 billion in OpenAI earlier and obtained an exclusive license to use it.