Gemini's new capabilities may not be far off. An analysis of Google app version 16.6.23 has revealed intriguing references to a mysterious “videogen”, a previously unknown term mentioned alongside “robin”, Gemini's internal codename. Messages such as “Work in progress…” and “We'll let you know when it's ready” suggest that a video generation feature is in development.
Such source code leaks have helped foreshadow major tech announcements in the past. While not an official confirmation, these clues strongly suggest that Google is testing video generation technology integrated into Gemini.
Gemini and the rise of creative AI
Google's entry into video generation wouldn't be a surprise. The company already has tools like Google Vids, a platform that assists users with editing and storytelling, without generating standalone videos.
With Gemini, Google could take a major step forward by integrating technologies similar to those in Imagen 3, its image generation model, to produce realistic animated sequences from simple text instructions. If this avenue is confirmed, Gemini could become one of the most advanced digital assistants on the market, combining text, image, and now video generation in a single tool.
An all-in-one digital assistant?
Gemini was designed to understand context and interact intelligently with its digital environment. The addition of video generation would reinforce this ambition by allowing users to create multimedia content without having to master complex software.
This development could have major implications for several sectors:
- Marketing and advertising: rapid creation of animated promotional content.
- Education: generation of interactive visuals in seconds.
- Social networks: production of personalized videos directly from an AI assistant.
With this approach, Google could position itself as a direct competitor to OpenAI's Sora, which has already demonstrated impressive capabilities in AI-based video generation.
An imminent release or a project in the making?
If Google is indeed working on integrating video into Gemini, no launch date has yet been leaked. It is likely that the technology is still in the internal testing phase and will only be unveiled once it is sufficiently mature.
The challenge will be twofold:
- Offer credible and usable video quality from the first versions.
- Ensure that the tool meets strict security and ethical criteria to prevent abuse and malicious use.
Google is making progress in multimodal AI
The potential arrival of the video generation in Gemini confirms that Google is pushing its AI assistant towards a multimodal approach capable of integrating text, images, and videos in the same environment. A logical development as the competition – OpenAI in the lead – is also accelerating in this area.
We will now have to wait for an official announcement to know how far Google is prepared to go with this technology and what its concrete applications will be for the general public.
0 Comments