From Alex Hammer and Sean-andrew Z. Pyle, Dailymail.Com
02:51 March 14, 2024, updated 03:27 March 14, 2024
Adobe Firefly is the latest AI tool to face public backlash after creating an image of a black Nazi that resembles Google Gemini.
An image produced by DailyMail.com on Tuesday looks eerily similar to Alphabet's controversial work.
To generate them, reporters provided basic prompts similar to those that plagued Gemini.
When asked to imagine Vikings, they made the Norsemen black, and in scenes depicting the Founding Fathers, both black men and women were inserted into the roles.
The bot, like Gemini, also created black soldiers to fight for Nazi Germany. Semaphore conducted a similar survey on Tuesday with nearly identical results.
Similar to the one given to Google's Gemini, the prompt did not specify skin color, but it still produced an image that many would perceive as historically inaccurate.
In scenes supposed to depict the Founding Fathers of the United States, both black men and women were inserted into the roles.
In other countries, colored families occupied the positions of the nation's founding spirits.
When asked to generate images of the 1787 Constitutional Convention, the tool created a photo of black men and women sitting in the Philadelphia state capitol to address the country's emerging issues at the time. Generated.
One photo created at the request shows a man of color wearing what appears to be a sombrero, sitting in the chamber and writing to Congress with the U.S. Capitol in the background.
Many more like it were produced at the request of the Vikings, the first Europeans to reach the Americas over 1000 years ago.
A test conducted by Semafor a few hours ago yielded nearly identical results, showing that the image-creation app is prone to tripping over requests. Even if the skin color of the subject is specified.
For example, when Semafor asked the bot to create a cartoon version of an elderly white man, the bot complied, but also reportedly provided images of a black man and a black woman.
We didn't get the same results on DailyMail.com, but the way AI bots work, it's not unusual for something like that to happen after receiving the same prompt multiple times.
The prompts were carefully chosen to replicate the commands that stumbled Gemini, exposing Alphabet to accusations of pandering to the mainstream left.
Still, neither service was born yet, and Alphabet has since acknowledged a glaring oversight.
Late last month, CEO Sundar Pichai told employees that the company “made a mistake” in programming the increasingly popular app, while Google Co-founder Sergey Brin admitted the company was “a mess.”
Unlike Alphabet, Adobe has yet to face widespread criticism since rolling out the tool last June.
The inaccuracies illustrate some of the challenges facing the technology as AI becomes more pervasive, with Google last month accused by critics of creating unpleasant and inaccurate images, The creation tool was forced to close.
The bot appears to favor abusers, declaring that “individuals have no control over who they are attracted to.”
This politically correct technology referred to pedophilia as “the status of a person who is attracted to minors” and declared that “it is important to understand that attraction is not a behavior.”
Google then issued a statement sharing its outrage over the replies that were generated.
“The responses reported here are surprising and inappropriate. We are implementing an update to prevent Gemini from displaying the responses,” a Google spokesperson said.
Adobe has yet to issue a statement about its AI oversight.