Microsoft is about to launch GPT-4, the AI update that could make ChatGPT able to create videos from simple text messages. Andreas Brown, CTO of Microsoft Germany, announced the news during the "AI in Focus - Digital Kickoff" event. GPT-4 will be "multimode" and will allow intelligence to translate text into other types of media. Braun explained that the new update will offer "completely different possibilities, for example video". Frankly, I also expect images and sounds: probably music too.
Not the first, but the best
Among the experiments in the field of AI that have come under the spotlight, there are already two systems capable of producing videos starting from text instructions. It's about Make A Video, AI owned by Facebook Meta (we talked about it here), And Dreamix, developed by Google (we talked about it here). However, it seems that the next version of ChatGPT powered by GPT-4 will be able to do more.
Microsoft has provided an example of use: the scenario is that of a call center. In the model illustrated by Braun, AI automatically converts phone conversations into text, saving employees time and effort. It is a simple taste: the characteristics of GPT-4 are all to be discovered and evaluated, and we will do so very soon. When? Starting as early as next week, the Redmond giant announced.
GPT-4, another page to turn
Although recent developments are getting us used to collecting news rather quickly, things are getting faster and faster. GPT-4 is just around the corner, and further developments of this technology are already being foreseen. AI will understand natural language with an even greater level of complexity and sophistication. They will be able to process and interpret information in an increasingly precise and contextual way, adapting quickly and responding to "dynamic" situations.
Most importantly, it will learn faster: the ability to "outsource" across a multimodal complex will add more "senses" to AI. On the horizon is the integration of these systems with other advanced technologies such as computer vision, or the robotics. We won't have "sentient" robots, but they will be (and that's for sure) vastly more advanced than any we've seen around since robotics started churning out attempts.
What can I say: I can't wait to "get my teeth on" GPT-4 as well. And you?