"Our research results have challenged the classical theories in the field of astrophysics!" On January 19, Nature published a major result of the cooperation between Liu Chao, a researcher at the National Astronomical Observatory of the Chinese Academy of Sciences, and researchers from Beijing Normal University, Nanjing University, Zijinshan Observatory and other institutions. It is believed that the "distribution law of initial mass of stars" will change significantly with the content and age of metal elements in stars, which challenges the classical theory of "the distribution law of initial mass of stars remains unchanged" and refreshes human's understanding of this basic concept. The international authority in the field of star initial mass function Professor Pavel Kulupa of the University of Bonn, Germany, commented: "This study, based on the high-quality data obtained from large sample observations, reveals that the initial mass function of stars in the Milky Way is related to the evolution history and environment of the Milky Way. This is very important for the in-depth understanding of the nature of star formation in different environments and at different times in the Milky Way." It has solved the 30-year dispute that new stars are always born in the vast universe, Many new stars of different masses will be formed in batches in the same star-forming region. The distribution law of the initial mass of stars is called "the initial mass function of stars" in astronomy. It describes the proportion of stars with different masses when a group of stars are born. "The initial mass function of stars is universal and unchangeable in all parts of the universe and at all stages of evolution." Since the concept was put forward in 1955, as a basic hypothesis, it has been widely used in many astrophysical research fields, such as the formation and evolution of galaxies, the structure and evolution of star clusters, the evolution of binary stars, even extrasolar planets and gravitational waves, and has become a "classical hypothesis" in astrophysical textbooks. "Some signs show that in the environment of active star formation, the proportion of massive stars is higher, which means that the initial mass function of stars may not be universal," said Zhang Zhiyu, a professor of astronomy at Nanjing University and a collaborator of the paper. However, due to the lack of direct and strong observational evidence, this dispute has existed for 30 years. "What troubled us at that time was that the number of stars near the sun that could be used for analysis was too small, and the statistical uncertainty was very high. The interpretation of the key signals of the extragalactic system of the Milky Way strongly depended on the physical model, which might be affected by the system deviation," Liu Chao said. The original results published this time give full play to the advantages of China's major scientific and technological infrastructure (LAMOST) hyperspectral data samples, and combine the European Space Agency Gaia satellite data. LAMOST currently holds the world record of the largest sample star spectral data. Gaia's performance is just complementary to that of LAMOST. The combination of the two is very helpful for star classification and can pick out "old" stars. It is this advantage that the research team was able to "select" 90000 star samples within 1000 light-years from the sun. While ensuring that the error is small enough, the dependence on the physical model is minimized. After the improvement of observation methods, Liu Chao's team first used the most intuitive star counting method, namely "counting stars", to make statistics on stars with different metal element contents and ages, and directly obtained the initial mass function of stars that are almost independent of any model from the observation point of view