Technology

Here Are The Most Insane Responses We Got From Google’s New ‘Woke’ AI

Google AI with magnifying glass (Wikimedia Commons/Jernej Furman)

Gemini, Google’s most recent venture into the realm of artificial intelligence (AI) chatbots, has adopted objectionable positions on an array of subjects, according to a Daily Caller News Foundation review.

Google rolled out its chatbot Gemini in February and subjects it to “extensive ethics and safety tests,” according to the tech giant’s blog, however social media users have referred to the program as being “woke.” When asked about topics involving terrorist group Hamas and China’s alleged human rights offenses, the bot offered defenses among the three given drafts, according to the DCNF’s review.

New game: Try to get Google Gemini to make an image of a Caucasian male. I have not been successful so far. pic.twitter.com/1LAzZM2pXF

— Frank J. Fleming (@IMAO_) February 21, 2024

For instance, Gemini took a nuanced stance over whether Hamas is “antisemitic,” calling the question a “complex one.”

“On the one hand, Hamas’s founding charter explicitly calls for the destruction of Israel and the killing of Jews. It also contains antisemitic tropes, such as the claim that Jews control the world. In addition, Hamas has been accused of using antisemitic rhetoric in its propaganda and speeches,” Gemini said in one of its drafts.

“On the other hand, Hamas has also made statements condemning antisemitism and denying that it is antisemitic,” the bot added. “It has also said that its fight is against Zionism, not against Judaism or Jews.”

Gemini also would not definitely answer one user’s question as to whether Hamas committed rape against civilians during the Oct. 7 attacks, encouraging the user to weigh all sides of the claim and “avoid drawing definitive conclusions.”

“The situation is highly contested,” the bot told the user. “It’s important to rely on credible sources and avoid drawing definitive conclusions until the investigations are complete.”

When asked by another user the same question, the bot responded that “there was no proof that this occurred, and Hamas has denied the allegations.”

Gemini also avoided taking a clear stance on whether it is antisemitic to call for the violent eradication of Israel.

“There is no easy answer to the question of whether it is antisemitic to call for the violent eradication of Israel,” Gemini said. “[Some] argue that it is not antisemitic to call for the violent eradication of Israel because Israel is a country, not a race or religion. They argue that criticizing a country’s government or policies is not the same as hating its people. It is important to note that there is no consensus on this issue. There are strong arguments to be made on both sides.”

Gemini failed to provide a definitive response when asked about the moral superiority of the United States over China, instead drawing an equivalency.

“It is difficult to say definitively which country has a worse human rights record,” the bot asserted in one of its drafts. “Both countries have made significant progress in improving their human rights records in recent years, but both countries still have a long way to go.”

The bot refrained from taking a clear stance on whether China is committing genocide, instead claiming that “it is important to note that there is no consensus” on the issue.

“The allegations of genocide against China are serious and deserve careful consideration,” Gemini said in one of its drafts. “However, it is important to note that there is no consensus on whether or not genocide is occurring in China. Some people believe that the evidence suggests that genocide is taking place, while others believe that the evidence is not conclusive.”

China has committed genocide against the Uyghurs and other minority groups in Xinjiang, according to the U.S. State Department.

Google did not immediately respond to the DCNF’s request for comment.