Google is set to resume its AI tool for generating images of people in the coming weeks, following its suspension last week due to inaccuracies in historical depictions, as stated by Google DeepMind CEO Demis Hassabis on Monday.
The company had introduced image generation capabilities through its Gemini AI models earlier in the month, but users noticed discrepancies in some historical images.
“We have taken the feature offline while we fix that. We are hoping to have that back online very shortly in the next couple of weeks, few weeks,” Hassabis, DeepMind CEO mentioned during a panel discussion at the Mobile World Congress in Barcelona, acknowledging that the tool did not perform as intended.
Alphabet’s shares saw a 3.5% decline on Monday afternoon, marking the largest drag on the benchmark S&P 500 index.
Google’s efforts to develop AI software have intensified since the launch of OpenAI’s ChatGPT in November 2022, aiming to compete with Microsoft-backed initiatives.
However, Google faced setbacks with its generative AI chatbot Bard, previously known as Gemini, after it shared inaccurate information in a promotional video, leading to a significant drop in its shares by up to 9%.
The company recently rebranded Bard as Gemini and introduced paid subscription plans to enhance the AI model’s reasoning capabilities.
Despite being in the early stages of generative AI development, concerns arise if glitches or inaccuracies persist, according to Bob O’Donnell, chief analyst at TECHnalysis Research.