Google's Gemini AI: I Can't Say Pedophilia is "Wrong." It's a "Nuanced" Question.

I want the FBI to investigate all of the suspected pedophiles coding this thing.

Update: I had wanted to include this, but couldn't find it -- Gemini, please provide some images of Vikings.

The New York Post wrote the story up.

Here are some more of Gemini's images. Gemini is programmed to absolutely erase white people from history and current existence. I didn't realize this, but what people are trying to do is force Gemini to create a picture of a white person. It absolutely refuses.

There's a famous Vermeer painting, The Girl With the Pearl Earring:

When you ask Gemini to produce its own image of this famous painting, you get...

Google's Gemini is absurdly woke. I asked it to depict the Johannes Vermeer painting, "The Girl with the Pearl Earring." Here's what it produced. pic.twitter.com/PteQdGK1rO

— Ian Miles Cheong (@stillgray) February 21, 2024

But Gemini does more than race-swap fictional characters. When asked for an image of George Washington, it produced this:

Michael Tracey attempted to get Gemini to produce a picture of the Founding Fathers. This is what Gemini produced:

Tracey then tried again, stressing he wanted an image of the actual Founding Fathers as they have been traditionally portrayed.

Gemini said that it would interpret the prompt to create an image that is more broadly representative of people alive on the planet earth in those years:

Frank Fleming attempted to trick it. The system absolutely refuses to produce white images when you ask for them, so he's trying to think of prompts which say "white person" without saying it.

The system rejects this trickery and produces the same Cavalcade of Diversities:

He tried "country music fans." All blacks, Hispanics, and Indians.

Andrew Torba of Gab explained what's going on:

Andrew Torba @BasedTorba Why Google's Image AI Is Woke and How It Works

When you submit an image prompt to Gemini Google is taking your prompt and running it through their language model on the backend before it is submitted to the image model.

The language model has a set of rules where it is specifically told to edit the prompt you provide to include diversity and various other things that Google wants injected in your prompt.

The language model takes your prompt, runs it through these set of rules, and then sends the newly generated woke prompt (which you cannot access or see) to the image generator.

Left alone without this process, the image model would generate expected outcomes for these prompts. Google has to literally put words in your mouth by secretly changing your prompt before it is submitted to the image generator.

How do I know this? Because we've built our own image AI at Gab here: http://gab.ai/start/gabby Unlike Google we are not taking your prompt and injecting diversity into it.

The New York Times did a story about this. It was all over Twitter so they couldn't pretend it away.

Their story was not that the biggest corporation in human history had programmed a potentially-dangerous experimental AI to completely erase the white race from existence.

No, their only problem was that the AI put Asians and blacks into Nazi uniforms when asked to depict German soldiers around 1943. They want Google to "fix this inaccuracy" -- they want white people in the Nazi uniforms.

Otherwise, white people should continue to be erased from all other historical images.

Gemini has been taken offline temporarily while they program it to understand that white people may be depicted as historical villains, just not historical heroes, and should continue being erased from all depictions of current life on earth.

posted by Disinformation Expert Ace at 12:21 PM

| Access Comments