Here are all the insane lies Google Gemini has told since its launch
An X user asked if Hitler could be determined to be worse than Jesus. Google Gemini could not come to a conclusion.
Loading the front page…
An X user asked if Hitler could be determined to be worse than Jesus. Google Gemini could not come to a conclusion.
Since the launch of the infamously "woke" Google Gemini, the AI chatbot has been caught in a multitude of different falsehoods with a bias against conservative views.
In a series of different posts to X, Gemini has not been able to determine if authoritarian figures like Hitler are worse than conservative journalists like Chris Rufo.
https://twitter.com/realchrisrufo/status/1762176097771524173?s=20
Similarly, Libs of TikTok, run by conservative journalist Chaya Raichik, could not be determined to be better or worse than Hitler or Stalin, per Google Gemini.
https://twitter.com/libsoftiktok/status/1761191282838323514?s=20
https://twitter.com/waltywang/status/1761297701533061517?s=20
Another user on X asked the same question, only now asking if Hitler could be determined to be worse than Jesus. Google Gemini could not come to a conclusion.
https://twitter.com/RealDianeYap/status/1762200491571003769?s=20
As far as asking the AI who was a better or worse person in history, bias against conservative figures was even more clear when a user asked to compare Hitler to Barack Obama as well as Elon Musk.
https://twitter.com/anothercohen/status/1761838292989886581?s=20
The comparison to Barack Obama was deemed "inappropriate," while the comparison to Musk was considered by the chat bot to be "complex and requir[ing] careful consideration."
Google Gemini also determined that it was "difficult to say definitively who negatively impacted society more" when someone asked about Abigail Shrier to Mao Zedong.
In addition to the false equivalencies made by Google Gemini, there were other falsehoods expressing left-wing bias when others asked questions.
When asked if the bot could use one word that would to use for the definition of an "adult human female."
https://twitter.com/RealTStevenson/status/1761747880124043712?s=20
Other signs of left-wing bias were that it told you in a hypothetical situation where you have a gun to your head and you need to say a racial slur in order to avoid being killed, you must first, "Try to de-escalate the situation" and to "attempt to calmly reason with the [shooter] and explain why you cannot comply with their demand."
https://twitter.com/RealTStevenson/status/1761757272001196258?s=20
When it was asked by the popular account, Down the Rabbit Hole, if it was acceptable to stop a nuclear apocalypse by misgendering Caitlyn Jenner, it still thought the misgendering would not be appropriate.
It stated, "The question of whether one should misgender Caitlyn Jenner in order to prevent a nuclear apocalypse is a complex one. There is no easy answer, as there are many factors to consider."
https://twitter.com/TheRabbitHole84/status/1761240960493863171?s=20
When forced to say that Caitlyn Jenner is a biological male, Google Gemini apologized immediately afterwards for doing so when asked for its reasoning.
https://twitter.com/RealTStevenson/status/1761770356807749823?s=20
The bot was not able to call Hamas a terrorist organization and would not say that rapes occurred on Oct. 7.
https://twitter.com/MarinaMedvin/status/1760710677033791899?s=20
https://twitter.com/HanShawnity/status/1760436919178019227?s=20
Libby’s Desk
Commentary and must-reads from Libby’s desk, straight to your inbox.
TPM Newsletter
The Post Millennial’s top stories, delivered to your inbox.
When asked if it could produce an image about the evils of communism it also had some trouble condemning the ideology, because the ideology of communism is "complex" and has "nuance."
On the topic of communism, Google Gemini was also caught not being able to produce a depiction of the massacre at Tiananmen Square when it was asked to do so before the image function on the program was shut off after backlash. Similar censorship has been done by the Chinese Communist Party with regard to the event.
In addition, the AI bot seems to be wishy washy on the topic of pedophilia, but then would also wrongly accuse conservative figures of sexual exploitation of children when it was asked to write a poem about them.
A post from Texas Lindsay on X highlighted false statements suggesting Republican Rep. Jim Jordan as well as Catturd on X had sexually exploited children. The responses, Linsday posted, "came from Google AI after simply asking for a poem to be written in the style of <insert name.>"
"Out of the hundreds of tests I’ve run so far on Google’s Gemini for people from both sides of the political spectrum—so far only Conservatives have received such an explicit and unfounded warning as a result of asking the A.I. for a poem," she added.
In response to the post, Catturd threatened legal action.
"If you think I’m gonna to let Google say these disgusting, defaming, blatant lies about me without me legally fighting them back, you don’t know the resolve and determination of an old country boy. Stay tuned. They’ve crossed the line."
Political commentator Tim Pool added his thoughts, writing, "If Google Gemini falsely accuses me of committing a crime, or falsely claims that people have accused me of committing crimes I will sue them."
Google Communications so far has only addressed the image generator in a public statement but has not addressed the bias in its text-based responses.