Mathematical simplicity bias and exponential functions — LessWrong