24.2 C
New York

Google Gemini: CEO Sundar Pichai admits ‘we got it wrong’


The CEO of Google, Sundar Pichai, has communicated through an internal memo to the staff, emphasizing that the errors in Gemini are “completely unacceptable” after the artificial intelligence (AI) app caused controversy.

The technology giant paused the image creation function on Gemini recently due to generating offensive and historically inaccurate results.

Instances included the technology depicting non-White founding fathers of the United States, mixed-race Nazi-era German soldiers, and incorrect race portrayals of Google’s co-founders. A specific concern was the reluctance of the chatbot to determine if Elon Musk posting memes is worse than Adolf Hitler causing mass casualties.

Pichai took a firm stance due to the severity of the incident, taking responsibility for the outcomes of the AI technology.

There was no evasion or denial regarding the incident. Instead, the 51-year-old Indian-American executive pledged to investigate the issues with Gemini, deploying experts to focus on the large language model (LLM).

“I acknowledge that some of its responses have offended our users and shown bias – to clarify, that’s completely unacceptable and we made an error,” Pichai stated.

“Our teams have been diligently working to address these concerns. We are already witnessing significant improvements across various prompts,” he added.

Critics have criticized Google, citing accusations of anti-White bias.

Google’s Response to the Gemini Controversy

AI is designed with safeguards and prompts to help prevent abuses and bias, but this recent incident highlights the risks of rapidly advancing technology. It underscores the importance of implementing effective safeguards to build trust and credibility.

Previously, OpenAI faced scrutiny for predominantly showcasing white individuals in professional roles while depicting black individuals in more menial positions.

Pichai mentioned that Google will work to rectify Gemini’s image generation in the upcoming weeks, without specifying a timeline for how long they will address this ongoing and evolving issue.

He mentioned that the company has been working tirelessly on this matter, stating, “No AI is flawless, especially in this nascent stage of the industry’s progress, but we understand the high standard expected of us and will persevere for as long as necessary.”

Image: Created by Midjourney

Related articles

Recent articles