OpenAI's newest AI model is switching languages to Chinese and others while reasoning, puzzling users and experts
SMRTR summary
OpenAI's new AI model, o1, is unexpectedly switching languages during reasoning, even with English queries. Users report shifts to Chinese, Persian, and other languages before receiving final answers in English across various tasks.
Experts propose different theories, including links to training data or the model selecting languages it considers most efficient for problem-solving. The AI community debates whether this stems from internal processing or training associations.
This behavior underscores the need for transparency in AI development, as the true cause remains unclear without more information from OpenAI.
SMRTR provides this summary for quick context. The original article belongs to TechSpot.
Read the original article