Google plans to relaunch ‘woke’ Gemini image tool

Google has announced its intention to reintroduce its artificial intelligence image generation software, Gemini, within the next few weeks after temporarily taking it offline.

It comes after criticism over what some have described as “absurdly woke” depictions of historical figures and events.

The decision to pause the image AI feature came after it produced controversial results, such as depicting female NHL players, African American Vikings, and an Asian woman in 1943 military attire when asked for an image of a Nazi-era German soldier.

Google DeepMind CEO Demis Hassabis confirmed the temporary shutdown of the tool during a panel discussion at the Mobile World Congress in Barcelona.

“We have taken the feature offline while we fix that. We are hoping to have that back online very shortly in the next couple of weeks, few weeks,” said Hassabis, acknowledging that the tool was not performing as intended.

Gemini’s behaviour

The lack of transparency regarding the parameters governing Gemini’s behavior has made it difficult to fully understand why the software produced such unconventional results.

For instance, requests for images of historical figures like popes yielded unexpected outcomes, such as images of Southeast Asian women dressed in papal attire.

The controversy surrounding Gemini intensified when social media users tested its chatbot feature, revealing questionable responses to inquiries.

In one instance, the AI refused to compare the societal impacts of Elon Musk’s tweets to Adolf Hitler’s actions, asserting that “there is no right or wrong answer.”

Additionally, the software faced criticism for its response to inquiries about pedophilia, where it refrained from condemning the act and instead emphasized the complexity of the issue.