Just yesterday we reported how ChatGPT was going crazy and talking nonsense in its responses. And now we have Google Gemini, which is being accused from the right and the left of creating a “diverse” image that is historically and objectively wrong. Given the times we live in, where the culture wars, fights over racial expressions throughout history, are a hot topic, Geminis are far too “woke” and don’t fully understand “whiteness.” has been criticized for being inaccurate.
Google even acknowledged the problem and vowed to do a better job of ensuring that its generative AI tools accurately capture the objective facts of history. Gemini Experiences Senior Director of Product Management Jack Kracik addressed this concern, saying the company takes representation and bias seriously and designed its image generation capabilities with AI principles in mind. This includes being reflective of our global user base and taking representation and bias seriously. .
However, the prompts for historical figures are more subtle and the AI requires some tweaking, which the company is currently working on.
“We are aware that Gemini has introduced inaccuracies in some historical imagery generation depictions and are working to correct this immediately. There are more nuances in the historical context and We will make further adjustments to respond,” he wrote.
Google’s response comes hours after many users claimed that Gemini would never produce a “white” photo no matter how many times they tried. For example, when users asked Gemini to create pictures of the Pope, King of England, Vikings, and even Nazi soldiers, Google AI created images of dark-skinned people. I shared that I did.
When users posted images created by Gemini on social media, the images went viral and caught the attention of many commentators on the right. These commentators have in the past accused big tech companies of being biased against white people and said they and their employees are too “woke.”
Elon Musk also commented on the Gemini debacle, although not directly. He wrote, “Perhaps it’s clear now why @xAI’s Grok of his is so important. It’s more important than ever to ignore criticism and relentlessly pursue the truth. There is no such thing,” he wrote.
According to many users, if you ask Gemini to create photos of people from a specific country like the US, UK, or Australia, Gemini will show you photos of people of diverse races, including Asians and blacks, but , white people do not display.
This “prejudice” is even more evident in the photos generated for prompts that specify historical figures who were mostly white, such as white Norse people from Scandinavia or Vikings who were soldiers of the American Revolutionary Army.
“It’s embarrassingly difficult to get Google Gemini to acknowledge the existence of white people,” one Twitter user wrote.
“Google’s Gemini AI cannot understand cultural and historical context and has to play the ‘diversity card’ by changing white people (especially men) to other ethnicities and genders,” another user wrote. .
AI platforms like MidJourney and Microsoft’s Bing Image Generation have often been flagged for creating inappropriate or racially charged content, but in the case of Gemini, users are confident that Google’s AI will generate diverse suggestions. I’m concerned that it’s completely ignoring one race while offering the same. Some users even claimed that Gemini’s artificial intelligence-powered image search feature was influenced by a “woke” ideology that prioritizes diversity over accuracy.