Google CEO Sundar Pichai expressed disappointment over the recent malfunctioning of the company’s Gemini AI chatbot, deeming the errors and biases in its responses as unacceptable. He addressed the internal memo to employees, acknowledging the offensive and historically inaccurate text and image generation by Gemini.
The Gemini team is putting in extra effort to rectify the issues and has already witnessed improvements, as reported by Pichai. Despite working tirelessly to enhance the chatbot’s performance, Gemini faced criticism for generating misleading images, such as Black Vikings and neglecting images of white individuals.
To address the public backlash, Google temporarily halted Gemini’s ability to create people-centric images and issued an apology for the erroneous outputs. While the company aims to reintroduce the image generator in the coming weeks, Pichai emphasized the need for structural changes to prevent similar incidents from occurring in the future.
The sentiment within Google regarding Gemini’s missteps varied, with some employees expressing embarrassment over the AI’s outputs. Pichai reassured the team’s commitment to delivering accurate and unbiased information across all products, particularly the evolving AI technologies.
Despite acknowledging that AI technology is still evolving and imperfect, Pichai reiterated Google’s dedication to meeting high standards and rectifying any shortcomings promptly. The company is set to investigate the root causes of the malfunction and implement solutions to prevent recurrence on a large scale.
