Google DeepMind CEO Demis Hassabis announces relaunch of Gemini AI image tool in upcoming weeks.
Google is set to reintroduce its AI tool for generating images of people in the coming weeks, according to Google DeepMind CEO Demis Hassabis. The tool was temporarily halted last week due to inaccuracies in certain historical representations.
Alphabet’s Google started creating images through its Gemini AI models earlier this month. However, some users reported on social media that it created historical images that were sometimes inaccurate.
“We’ve taken the feature offline while we fix it. We hope to bring it back online shortly in the next few weeks,” Hassabis said at a panel at Mobile World Congress in Barcelona.
The tool “didn’t work the way we intended,” he added.
Shares of Alphabet were down 3.5% on Monday afternoon, the biggest drag on the S&P 500.
Since the launch of OpenAI’s ChatGPT in November 2022, Google has been racing to produce AI software to compete with the Microsoft-backed company.
When Google launched its generative AI chatbot Bard a year ago, it had shared inaccurate information about images of a planet outside Earth’s solar system in a promotional video, causing its shares to drop by as much as 9%.
Bard was rebranded as Gemini earlier this month, and Google introduced paid subscription plans that users can choose to access the AI model’s enhanced reasoning capabilities.
“We’re in the early stages of generative AI development, but if glitches or inaccuracies continue, people will start to worry,” said Bob O’Donnell, principal analyst at TECHnalysis Research.
