Google Relaunches Gemini AI Image Creation Tool
Google has announced the relaunch of its Gemini AI tool, reinstating the controversial image generation function for some users. The move comes after the feature was temporarily suspended due to concerns over its ability to depict White individuals accurately.
The Alphabet-owned tech giant revealed the updates in a blog post on Wednesday, highlighting significant improvements to its text-to-image technology, Imagen 3. The updated version of Imagen 3 will soon be available to users subscribed to Gemini Advanced, Gemini Business, and Gemini Enterprise plans.
Google stated, “Over the coming days, we’ll also start to roll out the generation of images of people, with an early access version for our Gemini Advanced, Business, and Enterprise users, starting in English. We’ve worked to make technical improvements to the product, as well as improved evaluation sets, red-teaming exercises and clear product principles.”
Advanced AI Image Creation
Google’s Imagen 3 allows users to transform text prompts into visual images, enhancing creative expression with AI-generated art. In the latest demonstration, Google shared a sample prompt that instructed the AI to generate “an animated image of a tiny dragon hatching from an egg in a sunlit meadow, surrounded by curious glowing butterflies. Vibrant colors, detailed scales.”
Another prompt showcased the tool’s versatility, asking it to create an image of a “ball gown made of paper napkins in an elegant showroom.” The company emphasized that the updated tool comes with robust safeguards designed to ensure it adheres to Google’s design principles.
These safeguards limit the generation of inappropriate content, such as instructions for self-harm, explicit imagery, and excessively violent scenes. This overhaul follows backlash from the initial rollout of Gemini, where user prompts led to unintended and controversial depictions of historical figures and religious icons.
Addressing Past Issues
Previously, Gemini faced criticism for generating inaccurate and offensive representations, including images of World War II soldiers and religious figures that were inconsistent with historical and cultural expectations.
Some images controversially portrayed Nazi soldiers as Black and Asian and popes as female. The initial mishaps led CEO Sundar Pichai to label the problems with the AI app as “unacceptable.” Google has since acknowledged these missteps and assured users of the improvements made to prevent similar errors.
“We have made significant progress in providing a better user experience when generating images of people,” the company stated. They also emphasized that while the tool is still under development, the safeguards are aimed at preventing the creation of photorealistic depictions of identifiable individuals, minors, or excessively gory and sexual scenes.
Gradual Rollout and Future Plans
Despite the improvements, Google cautioned that the AI-based tool is still evolving. “Of course, as with any generative AI tool, not every image Gemini creates will be perfect, but we’ll continue to listen to feedback from early users as we keep improving. We’ll gradually roll this out, aiming to bring it to more users and languages soon.”
As of Thursday, the image-generation function remains inaccessible to all Gemini Advanced subscribers. When asked to generate images of people, the tool responded with, “Image generation of people is coming soon to Gemini Advanced.”
Originally posted on OpenDataScience.com
Read more data science articles on OpenDataScience.com, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Interested in attending an ODSC event? Learn more about our upcoming events here.
