Google CEO Sundar Pichai has sent an internal memo to staff, stressing the Gemini errors are “completely unacceptable” after the artificial intelligence (AI) app embroiled the company in controversy.
The tech giant suspended its image creation function on Gemini in recent days after it generated offensive and historically inaccurate results.
In some instances, the technology depicted non-White founding fathers of the United States, mixed-race Nazi-era German soldiers, as well as incorrect race portrayals of Google’s co-founders. A particular concern would have been the reticence of the chatbot to say if Elon Musk tweeting memes is worse than Adolf Hitler killing millions of people.
Pichai spoke out to present a stern position, given the seriousness of the incident, as he held his hands up for the output of the AI technology.
There was no intransigence or deflection over the incident, instead, the 51-year-old Indian-American executive vowed to get to the bottom of the issues with Gemini, after deploying specialists to focus their attention on the large language model (LLM).
“I know that some of its responses have offended our users and shown bias – to be clear, that’s completely unacceptable and we got it wrong,” Pichai said.
“Our teams have been working around the clock to address these issues. We’re already seeing a substantial improvement on a wide range of prompts,” he continued.
Critics have aimed fire at Google, including accusations of an anti-White bias.
1/ Google has halted its Gemini AI image generator due to historical and racial inaccuracies.
The AI was "missing the mark," per a company statement, after users came forward citing results like Nazi-era German soldiers being depicted as people of color.
Let's get into it. pic.twitter.com/X4Knqt7OZm
— Morning Brew ☕️ (@MorningBrew) February 23, 2024
How is Google responding to the Gemini controversy?
AI is built with safeguards and prompts included to supposedly mitigate abuses and bias, but this timely example shows the dangers of the rapidly emerging technology and why it is so important for effective safeguards to be put in place for trust and authority.
Previously, OpenAI was scrutinized for predominantly showing images of white people in professional positions whilst black people were depicted in typically menial roles.
Pichai indicated Google would be striving to fix Gemini’s image generation in the coming weeks, without putting an exact timescale on how long they would dedicate to what is an ongoing, evolving concern.
He stated the company has “been working around the clock” on this matter whilst highlighting “no Al is perfect, especially at this emerging stage of the industry’s development, but we know the bar is high for us and we will keep at it for however long it takes.”
Image: Created by Midjourney
