Google is working to fix its Gemini AI tool, CEO Sundar Pichai told employees in a memo on Tuesday, saying some of the responses to text and images generated by the model were “biased” and “totally unacceptable.”
Last week, the company suspended use of its tool that creates images of people after inaccuracies in some historical images it generated.
Pichai told officials that some of the tool’s responses offended users and showed bias.
“Our teams are working around the clock to address these issues. “We’re already seeing significant improvement across a wide range of prompts… And we’re going to review what happened and make sure we fix it at scale,” he said.
The company now plans to relaunch Gemini AI in the next few weeks. News website Semafor first reported the news, which was later confirmed by a Google spokesperson.
Since launching Microsoft OpenAI’s ChatGPT in November 2022, Alphabet-owned Google has been racing to create competing AI software.
It launched the generative AI chatbot Bard a year ago. Earlier this month, Google renamed it Gemini and launched paid subscription plans that users can choose from for better reasoning capabilities from the AI model.
Last week, social media platform Reddit struck a deal with Google to make its content available to train the search engine giant’s artificial intelligence models.
The contract with Alphabet-owned Google is worth around $60 million (roughly Rs 497 crore) per year, according to a source familiar with the matter.
The deal highlights how Reddit, which is gearing up for a high-profile stock market launch, is looking to generate new revenue amid fierce competition for ad dollars from the likes of TikTok and Facebook on the Meta Platform.
© Thomson Reuters 2024
For details on the latest launches and news from Samsung, Xiaomi, Realme, OnePlus, Oppo and more at Mobile World Congress in Barcelona, visit our MWC 2024 hub.
https://www.gadgets360.com/ai/news/google-fixing-gemini-ai-image-text-responses-unacceptable-ceo-sundar-pichai-5144334#rss-gadgets-all