Blog

by Zain Jaffer

Recently in February 2024, the search engine giant Google got criticized over its new Gemini AI image generation application. At the center of the criticism was Gemini’s inability to generate characters of white Caucasian people. Users who entered queries to generate the Pope, World War II German soldiers, George Washington, and others got inaccurate AI images of non-white characters for these people.

Google appeared to acknowledge these mistakes in a tweet by saying:

@Google_Comms
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.

I am obviously not Caucasian and find some Diversity, Equity, and Inclusion (DEI) measures reasonable. However many of these AI programs have gone overboard in their desire to not propagate stereotypes.

First of all, it is actually an insult to everyone to misrepresent the reality of history with woke utopian images of what we want. We all want to be treated fairly and equitably today, but we do not want to see incorrect images from history of people like George Washington, Abraham Lincoln, Winston Churchill, and the like. That actually misrepresents the objectives of DEI, some of which I do not agree with personally even if I am not Caucasian.

The part I really have a major problem with is how Google, one of the big bellwethers of the NASDAQ, passed this feature (or bug) off without some type of management review. This is not some test that requires expensive sophisticated equipment such as that of advanced chips to detect. It simply involves typing queries to see if the images generated are accurate.

This leads me to my next point which is the way these giant corporations not only have a monopoly on their grip of the total addressable market, but that they are using their power and influence to move societal values. Generally I have no issues with that, and we definitely need to reduce racism in this country.

The problem is when you change history to suit an agenda, you have crossed the line. Most people want accurate answers. They do not want altered images that advance an agenda, even if they themselves are victims of racism.

This is a lesson for companies and businesses that want to advance DEI to the world. Learn where to draw the line. While I may have used AI image generation as an example, there are other instances and situations where companies might overdo their DEI initiatives.

Consumers will eventually realize that they are not getting correct results or outcomes and will seek other applications from other companies and businesses to provide them what they want and need. That is the day that tech monopolies like Google (and similar providers) will start their decline.