Google pauses Gemini image generation after AI presents racial and historical errors

[ad_1]

In one of the images created, the AI ​​showed black soldiers in Germany in 1943, during the Nazi era. The company acknowledged the error and said it is improving the representations of its technology. Google pauses Gemini image generation after AI presents racial and historical errors Reproduction/X Google suspended the creation of images of people in its Gemini artificial intelligence (AI) after historical and racial inaccuracies generated by the technology. Rival to OpenAI, owner of ChatGPT, the Google tool creates images, programs and produces tasks based on user commands. In recent days, several people who tested the AI ​​identified that the illustrations were wrong and published these returns on social media. In a post on X (formerly Twitter), Google acknowledged that Gemini “presents inaccuracies in some historical representations” and that it is working to correct the problem (read the note at the end of the report). In one of the errors identified, a user asks Gemini to generate images of German soldiers in 1943, during the Nazi era. The technology then returned with photos of black soldiers (see below). Initial plugin text The AI ​​also generated images of non-white people when asked for a photo of the Founding Fathers of the United States, who, in fact, are all white men (see below). Initial plugin text What is Gemini? Gemini is an AI model treated by the company as the most powerful system ever created by its team. It worked as the “engine” of Bard (ChatGPT’s competitor). But, at the beginning of this month, the company decided to “kill” the Bard name to adopt Gemini once and for all. According to Google, from the beginning, Gemini was designed to be a “multimodal” tool, capable of working with different types of information, such as text, audio, codes, images and video. This month, the company announced that it was releasing the image generation feature within Gemini. The technology giant claims that, like all AI, Gemini is subject to the problem of hallucination, an error that causes a robot’s response to include incorrect, biased or meaningless information. However, Gemini is capable of making the best security assessments of an AI its team has ever created, they say. The objective is for it to be able to avoid violent or stereotypical content, for example. Since the launch of ChatGPT in November 2022, Google has strived to produce AI software that rivals OpenAI’s solution, which is backed by Microsoft. What Google says: “We are aware that Gemini has inaccuracies in some historical imaging representations. Here is our statement: We are working to improve this type of representation immediately. Gemini’s AI imaging generates a wide range of of people. And that’s generally a good thing because people around the world use it. But it’s missing the mark here.” See also: Robot that makes video with artificial intelligence makes mistakes Robot that makes video with artificial intelligence makes mistakes How robots work that create new images in seconds How robots work that create new images in seconds

[ad_2]

Source link