FAST DOWNLOAD
According to the company, it will be bringing its latest imaging model to Gemini Apps in all languages ”in the coming days.”
Google announced Wednesday that it has ” significantly improved ” the user experience when generating “photorealistic” images of people using its artificial intelligence (AI), Gemini, after inaccurate and biased responses from its previous model went viral.
The previous tool received a lot of criticism for, for example, having generated images of an Asian woman or a black man after being asked to illustrate a “German soldier from 1943”, a date in which Nazism was on the rise in the country and racial diversity was practically impossible.
To quell the controversy, the company removed Gemini’s ability to create any human.
According to a statement released today by Dave Citron, director of product management at Gemini Experiences, Image 3, Google’s new text-to-image conversion model, will be configured to not represent photorealistic images of public figures, minors or violent images.
“With Image 3, we’ve made significant progress in improving the user experience when generating images of people (…) Of course, as with any generative AI tool, not every image created by Gemini will be perfect,” Citron adds.
According to the company, it will be bringing its latest imaging model to Gemini Apps in all languages ”in the coming days.”
Personalized Assistants
The tech giant also said that Gemini Advanced, Business and Enterprise users will be able to start using Gems starting Wednesday, a new feature that allows users to customize Gemini to create AI “assistants” for specific topics.
“(The user) can customize Gems to act as experts on a topic or fine-tune them to meet their specific goals. Simply type in instructions for their Gem, give it a name, and chat whenever they want,” Citron said in the statement.
