The German branch of Microsoft has shared some information about the highly anticipated next version of OpenAI Artificial Intelligence. GPT-4 will be announced next week, and should be able to handle images and videos.
The general public discovered the existence of OpenAI’s GPT-3.5 language model with the publication of ChatGPTChatGPT late last November. The next version, GPT-4, was expected for this year, but the launch is now said to be imminent. At a German conference AI in focus around Artificial Intelligence, Andreas Braun, CTO of MicrosoftMicrosoft Germany, said GPT-4 is expected to be announced in the coming days.
« Next week we will present the GPT-4, which will offer multimodal models offering completely different possibilities, for example videos “, he said. The announcement should logically fall during Microsoft’s conference on the future of work with AI (Future of Work with AI) scheduled for Thursday, March 16 at 5 p.m. French time. He did not specify whether Microsoft or OpenAI will make the announcement. The two companies are partners, with Microsoft having invested heavily in OpenAI and integrating its Artificial Intelligence into Bing.
GPT-4 support video contents
This new version of GPT is announced as being multimodal, in other words able to understand and generate several types of information such as images, audio and videos, unlike GPT-3.5 (and therefore ChatGPT) which is limited to text . GPT-4 could integrate (or even replace) the firm’s other AIs, such as Dall-E, which generates realistic images from a sentence, or Whisper, the next-generation voice recognition.
However, it will probably take some time before you can use it. The announcement may only concern the GPT-4 model, not its integration into ChatGPT, Bing or a new interface. However, given the popularity of AI, OpenAI may face strong demand from the general public as well as businesses.
