Google announced the suspension of its artificial intelligence tool, capable of generating images of people, due to errors in certain historical representations created by the model. This setback is the latest challenge the company faces in its attempt to compete with OpenAI and Microsoft.
At the beginning of the month, Google introduced image creation functionality through its AI models, called Gemini. However, in recent days, it has been noted by users on social media that the model occasionally produces inaccurate historical representations.
America's Founding Fathers, Vikings, and the Pope according to Google AI: pic.twitter.com/lw4aIKLwkp
— End Wokeness (@EndWokeness) February 21, 2024
“We recognize that Gemini has presented flaws in some historical representations regarding image generation,” said Google last Wednesday.
Since the launch of ChatGPT by OpenAI in November 2022, Google has been dedicated to developing AI software that can compete with the innovations presented by the Microsoft-backed company.
A year ago, when launching its generative AI chatbot called Bard, Google published misleading information about photographs of a planet outside our solar system in promotional material, causing its share price to drop by up to 9%.
Bard was renamed Gemini earlier this month, and Google announced paid subscription plans so users could access the AI model's enhanced reasoning capabilities.
“Historical events are complex and we are committed to making the necessary adjustments to respect that complexity,” said Jack Krawczyk, senior director of Gemini products at Google, last Wednesday.