This is a short essay based on the author's tiny experience of cognitive science study. It argues that traditional methods of cognitive science such as production system, protocol analysis, questionnaire through crowd sourcing to name a few are not enough for unraveling detailed mechanism of cognition and thought processes in our brain. After checking pros and cons of these traditional methods, new methodologies are speculated. These new methodologies will use machine learning, the Internet, and robots to their maxima. They try to gather massive recordings of our daily sensory data plus our internal reactions. Recent rapid progress in sensor technologies and machine learning makes it possible to create (i) tools to gather an unprecedented amount of sensory data and (ii) robots which facilitate such massive data gathering.