Google made a mistake with the generation of images in the Gemini chatbot – Kommersant

One of the founders of Google Sergey Brin stated at a conference in San Francisco that the company made a number of errors with the generation of images in the Gemini chatbot. The company, he said, did not pay enough attention to testing, which is why the AI ​​bot could produce historically inaccurate images.

Previously, Gemini users began to complain about strange images of the chatbot, which is why it could be suspected of racism. After this, Google temporarily suspended the image generation function of Gemini, but users noticed AI errors in text responses. According to Sergey Brin, Google does not fully understand why Gemini behaves unethically in some cases, but he assures that this was not part of the company’s plans.

Sergey Brin left Google leadership in 2019, but in 2023 back to active work and began developing in the field of AI and chatbots. Now he is one of the main developers of the Gemini chatbot. According to Sergey Brin, he “sort of came out of retirement because of the exciting developments in AI.”

Evgeniy Fedunenko