I've been reading papers on MIT dspace recently and man... some of them were written by people when they were undergrads and they're being cited even today by top scientists in national labs and/or AI research labs. How do people know what they want to do and excel at it at such a young age?