Honestly, I think this whole thing is blown over. Gemini has a huge reach, so they played it safe by doing the same thing OpenAI has been doing (system instructions for their AI services like GPT-4) and tried to max out diversity. This decision is juicy bait for anyone who believes that America is in some culture war. This is unsurprising if you live in America and are subjected to extremely high levels of marketing and polarizing headlines.
Please note: I think it is clear that the Gemini models still need work to be historically representative and accurate, however, I do not think there is any real danger in airing with the side of caution and trying to fight the innate biases in the training data. It seems like this was Google's goal.
Please note: I think it is clear that the Gemini models still need work to be historically representative and accurate, however, I do not think there is any real danger in airing with the side of caution and trying to fight the innate biases in the training data. It seems like this was Google's goal.