Google announced its plans to relaunch its AI-powered image generation tool, Gemini, within the next few weeks. The tool was temporarily suspended in February 2024 after users flagged historical inaccuracies in some generated images.
Demis Hassabis, CEO of Google DeepMind, the company's AI research division, addressed the issue at the Mobile World Congress in Barcelona, Spain. He acknowledged that the tool "was not working the way we intended" and emphasized Google's commitment to fixing the identified problems.
The initial launch of Gemini earlier in February marked the introduction of image generation capabilities powered by Google's Gemini AI models. However, concerns arose when users on social media pointed out inconsistencies in depictions of historical figures and events. These inaccuracies, while not explicitly detailed by Hassabis, likely involved issues with bias, representation, or factual errors within the generated images.
Google's decision to take the tool offline demonstrates the company's effort to address responsible development and deployment of AI technology. The temporary suspension allowed them to address the identified shortcomings and implement necessary improvements.
Hassabis offered a timeframe for the relaunch, stating they are "hoping to have [the tool] back online very shortly in the next couple of weeks, few weeks." He did not disclose the specific nature of the fixes implemented but emphasized Google's ongoing commitment to responsible AI development.
The relaunch of Gemini comes amidst a growing landscape of AI-powered image generation tools. Competitors like OpenAI's Dall-E 2 have garnered significant attention, highlighting the increasing demand and potential of this technology.
However, concerns regarding potential misuse and ethical implications of such tools remain. The ability to generate realistic images, particularly those depicting historical figures or events, raises concerns about the potential for manipulation and the spread of misinformation.
Google's approach of addressing the identified issues in Gemini and implementing fixes before relaunch reflects the ongoing challenges and considerations in responsible AI development. As the field of AI image generation continues to evolve, addressing these challenges proactively will be crucial in ensuring the ethical and responsible use of this powerful technology.