Microsoft teased a mysterious AI event for March 16th, and it appears that GPT-4, which includes multimodal support, will be a significant ChatGPT upgrade this week. Given that ChatGPT only invaded the digital landscape three months ago and that we are still learning what it can accomplish and how it can disrupt tech as we know it, most people may not understand what that means.
A multimodal ChatGPT chatbot is a significant improvement over AI, which already responds to your questions in a human-like manner. At the moment, ChatGPT only allows text input or one type of communication. Text, audio, video, and image inputs are all supported by GPT-4. It is multimodal because of this, a quality that might greatly expand the AI’s potential.
Microsoft USA simply hinted at the March 16th event last week, with no information about ChatGPT’s GPT-4 upgrade. Microsoft Germany, however, took things a step further and essentially soft-launched GPT-4. According to Heise.de, the business held a presentation last week in Germany where it discussed the GPT-4 upgrade.
GPT-4 may be a unique feature of Microsoft’s Bing search engine, which already supports ChatGPT, or it may be a built-in improvement to ChatGPT. Microsoft Germany did confirm that GPT-4 will be multimodal and arrive this week.
Microsoft CTO Andreas Braun said, “We will launch GPT-4 next week. There, we will have multimodal models that will offer completely different possibilities, for example films.
The underlying technology, AI that understands natural language, was dubbed a “game changer” by Braun. And he disclosed that ChaGPT will enable multi-language support and all languages. It may be useful to ask something in German and then receive a response in Italian.
Another Microsoft Germany executive, Holger Kenn, added that a multimodal ChatGPT bot can translate text into pictures, audio, and video upon request.
How would multimodal technology in ChatGPT -4 benefit users?
Users should be able to use a variety of input formats to receive the answers they require, even though many aspects of GPT-4 remain unknown. By moving beyond words, AI may now watch videos on YouTube or listen to audio files before responding to queries.
Microsoft provided an illustration of how ChatGPT’s multimodality might benefit enterprises. After listening to the recorded support calls, the AI may automatically textually summarise the calls. A large Microsoft customer in the Netherlands that has to summarise 30,000 calls per day would save 500 work hours per day if this were implemented. For such a work, ChatGPT setup would only take a few hours.
Even with the GPT-4 upgrade, Microsoft warns that ChatGPT won’t always be dependable. Microsoft is developing confidence measures to increase the dependability of the chatbot.
How users will test GPT-4, though, and whether OpenAI will simply make it available within ChatGPT later this week are both unknown. Early in March, Microsoft secretly debuted Kosmos-1, a multimodal AI that accepts picture input. Microsoft is also hosting an AI-focused event on Thursday.
Microsoft may be one of the major investors in OpenAI technology, but OpenAI will keep improving its chatbot. And doing so entails making GPT-4 widely accessible.
