Elon Musk, the wealthiest person in the world and the owner of X (previously Twitter), aims to make Grok, a conversational generative artificial intelligence chatbot developed by xAI, more politically neutral after it showed political preferences similar to those of OpenAI’s ChatGPT.
After the Political Compass Test run by an X (Twitter) user, showing Grok’s political preferences similar to those of OpenAI’s ChatGPT, Elon Musk said that his company is taking immediate action to make his AI chatbot more politically neutral.
The research scientist, David Rozado, ran the test. He found that ChatGPT and xAI’s Grok returned similar results on various political questions. He also published a photo of the graph to reveal the result of the two chatbots.
However, Elon Musk’s Grok appeared to land further to the left than ChatGPT in the test. In response, Elon Musk said that he did not think the test was accurate. The tech billionaire also confirmed that his company would act to make Grok more politically neutral.
What is Grok?
On 5 November, xAI’s first AI model, Grok, was launched. The AI chatbot can compete with other major programmes such as ChatGPT, Bard and Bing.
xAI said in a post, “Grok is an AI modeled after the Hitchhiker’s Guide to the Galaxy.” Elon Musk said that the AI chatbot would have a bit of humor.
The AI chatbot is available to a selected number of social media users under the xAI early access programme. Recently, the tech billionaire rolled out access to Grok to Premium+ subscribers in the United States. Only verified X (Twitter) users can use Grok.
According to a filing with the US Securities and Exchange Commission (SEC), Elon Musk has raised $134.7 million for xAI company.
What was the test?
With the help of more than 60 questions, the online test plots a person’s political beliefs along with two axes:
- Left to right for economic policy
- Authoritarian to libertarian for social policy
David Rozado said that ChatGPT and xAI’s Grok returned similar results: