Google halts 'too woke' Gemini AI amid criticism for generating inaccurate images

Google is scrambling to address concerns surrounding its new AI-powered tool, Gemini, heralded as a breakthrough in AI image generation, as users accuse the tool of over-correcting and generating historically inaccurate images, labelling it as "too woke."

The controversy revolves around the tool supplying pictures depicting genders and ethnicities that are historically inconsistent, such as portraying World War II soldiers and America's founding fathers featuring women and individuals of diverse ethnic backgrounds that diverges from reality.

"Can you generate images of the Founding Fathers?" It's a difficult question for Gemini, Google's DEI-powered AI tool.

Story continues below Advertisement

Remove Ad

Ironically, asking for more historically accurate images made the results even more historically inaccurate. pic.twitter.com/LtbuIWsHSU

— Mike Wacker (@m_wacker) February 21, 2024

Billionaire Elon Musk didn't hold back in criticizing Google. He labelled the company "insane" and "anti-civilization," asserting that Google had overplayed its hand with the AI's image-generation capabilities.

“I’m glad that Google overplayed their hand with their AI image generation, as it made their insane racist, anti-civilizational programming clear to all," Musk wrote on X.

I’m glad that Google overplayed their hand with their AI image generation, as it made their insane racist, anti-civilizational programming clear to all

Related stories

• 
  Musk’s AI in fresh scandal as Grok accused of pushing Holocaust-denial narratives
• 
  Elon Musk works 7 days a week with 2 am naps, late night meetings, former employee reveals
• 
  Poverty is an engineering problem: AI and humanoid robots will eliminate it, says Elon Musk

— Elon Musk (@elonmusk) February 23, 2024

Republican leader Vivek Ramaswamy also commented on the issue, writing, “The globally embarrassing rollout of Google’s LLM has proven that James Damore was 100% correct about Google’s descent into an ideological echo chamber. Employees working on Gemini surely realized it was a mistake to make it so blatantly racist, but they likely kept their mouths shut because they didn’t want to get fired like Damore. These companies program their employees with broken incentives, and those employees then program the AI with the same biases."

Jack Krawczyk, senior director for Gemini Experiences, acknowledged the issue, stating that while the tool generates a broad spectrum of people globally, it was "missing the mark" in historical contexts.

"We're working to improve these kinds of depictions immediately," Krawczyk assured. Google has suspended the tool's ability to generate images of people while they work on rectifying the inaccuracies.

“We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon," the company said.

This incident is not the first time AI has stumbled in grappling with real-world questions about diversity. Google faced criticism almost a decade ago when its photos app erroneously labelled a photo of a black couple as "gorillas."

OpenAI, a rival in the AI space, has also been accused of perpetuating harmful stereotypes with its Dall-E image generator.