“LLMs are not mechanical tools that provide the same benefit for everyone. They are rather intellectual tools that interact with human culture and creativity. The influence is mutual where not only are the models affected by the data we train on, but also our culture and the data we generate will be influenced by LLMs,” said Rio Yokota, professor at the Global Scientific Information and Computing Center at the Tokyo Institute of Technology.