We model agents who get utility from their beliefs and therefore interpret information optimistically. They may exhibit several biases observed in psychological studies such as optimism, procrastination, confirmation bias, polarization, and the endowment effect. In some formulations, they...