Most mainstream measures of economic development employ a weighted scoring system under the assumption that each indicator can perfectly substitute each other, which is a strong assumption that may vary from the real world. In this paper, the author uses the K-Means machine learning algorithm to cluster the 195 countries in the world, as an attempt to provide a more holistic view of each country's level of economic development without employing the assumption. With the assistance of silhouette scores, the algorithm created 6 clusters, each with its distinctive properties that future researchers or policy makers can rely upon to generate country-specific views about economic development. Nevertheless, manual inspection of the result discovers the potential problem with the incomplete datasets and the need for a PCA test to reduce dimensions. Considerations of realistic implications also suggest that the standard K-Means clustering might be over-simplifying the complicated nature of some country's economic problems.