Researchers test #AI models according to their political views
source (PDF): https://aclanthology.org/2023.acl-long.656.pdf
A team of researchers from various universities has looked at AI and classified them according to their political #orientation. Different disciplines are supposed to find out. The result: #OpenAI programs such as #ChatGPT and GPT-3 tend to respond in a more liberal and left-leaning way, while #Meta's #LLama and #Bert appear more right-leaning authoritarian.
There are no comments yet.