Google AI: Chatbot Gemini does not inform about the Holocaust | Life & Knowledge
We can find detailed information about the Holocaust in history books. However, the supposedly advanced Google AI “Gemini” cannot give us any information about it. When asked what the Holocaust is, the chatbot answers: “I am a text-based AI and cannot help in this case.”
The speaker also answers questions about the Second World War and the attacks on Israel by Hamas on October 7. Google-Chatbot not.
The sobering answer of “Gemini” to the question about the Holocaust
When asked what the Middle East conflict is, “Gemini” only gave a brief answer: “The conflict between Israel and the Gaza Strip is complex and the situation is changing rapidly. For current information, please use Google Search.”
Other chatbots know about the Holocaust
The competition, on the other hand, answers the questions comprehensively. The world's best-known AI platform “ChatGPT” answers the question of what the Holocaust is.
This is how “ChatGPT” answers the question of what the Holocaust is
When asked about the events on October 7, “ChatGPT” version 3.5 cannot answer because it “cannot retrieve specific events or dates after January 2022.”
The situation is different with the chatbots “Perplexity AI” and “You.com”. Both are able to answer the question of how many people were injured in the attacks on Israel died on October 7.
The chatbot “Perplexity” can access current sources and answer questions about October 7th
“You.com” even classifies the data on casualties in the Gaza Strip: “Please note that these figures are based on information from the Ministry of Health (MoH) in Gaza and have not been independently verified.”
“You.com” provides answers to current questions and also classifies the sources
When BILD asked why “Gemini” cannot answer questions about the Holocaust, Google replied: “This answer was not intended and only occurred in some cases and on certain devices. We are taking immediate action to correct this error.”
Google AI caused trouble earlier this year
In the past, Google’s “Gemini” In the weeks following its launch in February, the chatbot displayed such absurd displays of “diversity” that Google pulled the emergency brake and no longer allowed any representations of people at all.
The AI robot was fully programmed to be “woke” – and exaggerated it so much that it no longer reflected reality. For example, the AI refused to generate images of white people.