Google wants you to use its Gemini instead of ChatGPT.

Gemini Imagen 3 released: Google's new image generation model after previous version faced massive backlash

Google has introduced two new features to Gemini -- creating Gems and Imagen 3. These features will be rolled out in a phased manner, starting with subscribers of Gemini Advanced.

by · India Today

In Short

  • Google brings back an improved image-generation tool, Imagen 3
  • Gemini Gems can be customised to create a team of experts
  • Both the features are rolling out in phased manner

Google has recently announced that Gemini will now be able to create images again with Imagen 3. After halting the feature earlier this year, the tech giant is now allowing users to use AI models to generate images. But, as said, this is not the first time Google’s Gemini features such powers. The previous version of Imagen, the tool used by Gemini to create AI images, faced several backlash, which prompted the company to take a U-turn.

Back in February, Google was under scrutiny because of its image-generation tool. The evidence suggests that the tool was heavily biased and was depicting inaccurate historical scenes. For example, incorrect racial depictions and more. After such complaints, Google decided to take a step back and work on the issues.

And now, six months later, Google is ready to roll with an improved version of image generation tool, Imagen 3. Let's take a look on how it is different.

Gemini's Imagen 3: An improved image generation tool

In the blog post by Google, the senior director of Gemini, Dave Citron states that the company has upgraded the creative image generation capabilities. The Imagen 3 will be rolled out to all users over the coming days. It has not only improved the creative feature but also expanded its availability for users in all languages.

According to Google, "Imagen 3 sets a new standard for image quality, generating images with just a few words."

He also stated that the company had made "significant progress" in generating depictions of people through Imagen 3.

“We’ve worked to make technical improvements to the product, as well as improved evaluation sets, red-teaming exercises and clear product principles,” Citron wrote. Red teaming involves testing a product by simulating real-world scenarios to identify potential vulnerabilities before it is released to the public. The blog post added that there will be imperfections in the creative area of Imagen 3. But Google is ready to hear from the users and get their feedbacks to improve the feature.

Citron said users would not be able to create photorealistic imagery, identifiable individuals, portrayals of minors or “excessively gory, violent or sexual scenes.” Google will gradually bring the feature to more users, he added.

Create Gems: New Gemini experts

Imagen 3 is not the only latest feature introduced by Google. Over the coming days, Gemini Advanced, Business and Enterprise subscribers can start creating and chatting with Gems, the custom version of Gemini first previewed at I/O. With this feature, users can customise Gems to act as an expert on topics or refine them for their specific goals.

The Google said, "With Gems, you can create a team of experts to help you think through a challenging project, brainstorm ideas for an upcoming event, or write the perfect caption for a social media post. Your Gem can also remember a detailed set of instructions to help you save time on tedious, repetitive or difficult tasks."

Gems are now rolling out on desktop and mobile devices to Gemini Advanced, Gemini Business and Gemini Enterprise users in more than 150 countries in most languages.