GPT-4 will be introduced next week, and it may let you create AI-generated videos
Next week could be a major step forward for AI technology. Andreas Braun, Chief Technology Officer at Microsoft Germany, recently confirmed that GPT-4 will be unveiled next week (beginning March 19, 2023). GPT-4 will offer "completely different possibilities" than the current GPT-3.5 model, including video.
Braun commented on GPT-4 at an event called AI in Focus – Digital Kickoff(opens in new tab) (via Heise). The following quote is a translation of Braun's remarks:
"We will introduce GPT-4 next week, where we have multimodal models that will offer completely different possibilities – for example, videos."
ChatGPT, which launched in November 2022, is powered by the GPT-3.5 model. The next major version of the Large Language Model (LLM), GPT-4, should be significantly more powerful. It will also have new capabilities.
At the moment, ChatGPT and other technology powered by GPT-3.5 is limited to text-based answers. Braun's comments suggest that could change with the rollout of GPT-4. The LLM's multimodal models could pave the way for video creation and other types of content.
Microsoft unveiled a new version of Bing powered by ChatGPT earlier this year. While early rumors speculated that the search engine may have been powered by GPT-4, that was not the case. Instead, Microsoft uses proprietary tech called Prometheus to generate answers quickly that rely on real-time information. The company detailed how Prometheus powers the new Bing last month.
Microsoft has invested heavily into OpenAI, the makers of ChatGPT. The companies have also worked closely together to improve AI capabilities. The fact that Braun discussed GPT-4 at a Microsoft event indicates how closely the two organizations are working together.
Windows Central Newsletter