Clustering similar thoughts in a big pool of data

Would love to get your feedback on whether this new way of looking at data is useful or not