GPT-4 coming soon, may bring video, photo inputs to ChatGPT and Bing AI

Microsoft and OpenAI are gearing up to release the next big update for the Large Language Model (LLM) powering Bing AI and ChatGPT.

Generative Pre-Trained Transformer (GPT) is the language model powering OpenAI's ChatGPT and Microsoft's ChatGPT-infused Bing AI.

The version powering the technology is GPT-3.5 but new reports suggest that a big update might be on the way soon.

A report by German publication Heise, quoted Microsoft Germany's Chief Technical Officer Andreas Braun as saying said that the technology giant is looking to release the GPT-4 update as early as next week.

Also Read | MC Explains | The difference between GPT-3 and GPT-4

Braun said that the new model is multimodal, which means it can operate using various sources of input—audio, video, text and more.

Microsoft Germany's Director of Business Strategy Holger Kenn said GPT-4 can help call centres by automatically converting phone conversations to text. It would also summarise the call after it ended.

Also Read | What is Visual ChatGPT, and what does it do?

Besides being multimodal, GPT-4 will also focus on streamlining the existing model and improve performance, which can lead to faster responses to queries.

Both OpenAI and Microsoft are tight-lipped on potential applications for GPT-4 but it is safe to assume that ChatGPT will be the first benefit from the new update, which, in turn, will also improve Bing AI.