Gemini, the ambitious AI project at Google, is already so far into development that it could launch this year.

Since ChatGPT’s rapid growth, AI research has seen a significant boost at Google and its various teams. Chatbot Bard was created by Google Brain and DeepMind, owned by Google for several years, even combines AI with robotics. Both teams were brought together for the joint project Gemini AI.

Initially it looked as if Google was planning to launch Gemini this year, but that has now been postponed until January 2024 at the earliest. The information reports that the AI ​​model still achieves an unsatisfactory score for non-English search queries. CEO Sundar Pichai is said to have decided that postponing the launch was the best option.

Written in the stars

We reported on Gemini back in June, when the project was still in full development. Meanwhile, the newly founded AI family of Brain and DeepMind at Google is not idle; Gemini has been and is being extensively trained in many areas. Even co-founder Sergey Brin is actively involved in the project. So it’s Google, man.

The intention is not just to compete with OpenAI, but to become much better and exploit AI’s capabilities even further. In addition to being a large language model, Gemini also needs to generate contextual AI images.

That is not all. The model should also be able to read flowcharts and give you the ability to control software with your voice. With the huge video inventory that Google has on YouTube, Gemini would even have videos (transcripts) in the training material. All within a legal framework.

According to the (anonymous) source within the project, Gemini will soon power products like Google Docs and developers will have to pay for access. After being used on various products, there will most likely also be an app.

This article originally appeared on August 17th. The text has been updated with the latest information.